From in%@vtcs1 Mon Nov 3 02:23:09 1986 Date: Mon, 3 Nov 86 02:23:05 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #240 Status: R AIList Digest Thursday, 30 Oct 1986 Volume 4 : Issue 240 Today's Topics: Queries - PD Parser for Simple English & Public Domain Prolog & Faculty Compensation & Hierarchical Constraints & Model-Based Reasoning & Monotonic Reasoning, Neural Networks - Simulation & Nature of Computation ---------------------------------------------------------------------- Date: 26 Oct 86 19:10:57 GMT From: trwrb!orion!heins@ucbvax.Berkeley.EDU (Michael Heins) Subject: Seeking PD parser for simple English sentences. I am looking for public domain software which I can use to help me parse simple English sentences into some kind of standardized representation. I guess what I am looking for would be a kind of sentence diagrammer which would not have to have any deep knowledge of the meanings of the nouns, verbs, adjectives, etc. The application is for a command interface to a computer, for use by novice users. C routines would be ideal. Also, references to published algorithms would be useful. Thanks in advance. -- ...!hplabs!sdcrdcf!trwrb!orion!heins We are a way for the universe to know itself. -- Carl Sagan ------------------------------ Date: 27 Oct 86 14:26:33 GMT From: ihnp4!drutx!mtuxo!mtune!mtunf!mtx5c!mtx5d!mtx5a!mtx5e!mtx5w!drv@ ucbvax.Berkeley.EDU Subject: NEED PUBLIC DOMAIN PROLOG A friend of mine needs a copy of a public domain Prolog that will run on a VAX 11/780 under Unix. If such a program exists, please contact me and I will help make arrangements to get it sent to him. Dennis R. Vogel AT&T Information Systems Middletown, NJ (201) 957-4951 ------------------------------ Date: Tue, 28 Oct 86 09:31 EST From: Norm Badler Subject: request for information If you are a faculty member or a researcher at a University, I would like to have a BRIEF response to the following question: Do you have an "incentive" or "reward" or "benefit" plan that returns to you some amount of your (external) research money for your own University discretionary use? If the answer is NO, that information would be useful. If YES, then a brief account would be appreciated. If you don't want to type much, send me your phone number and I will call you for the information. Thanks very much! Norm Badler Badler@cis.upenn.edu (215)898-5862 ------------------------------ Date: Mon, 27 Oct 86 11:47:52 EST From: Jim Hendler Subject: seeking info Subject: seeking info on multi-linear partial orderings I recently had a paper rejected from a conference that discussed, among other things, using a set of hierarchical networks for constraint propogation (i.e. propogating the information through several levels of network simultaneously). One of the reviewers said "they apply a fairly standard AI technique..." and I wonder about this. I thought I was up on various constraint propagation techniques, but wonder if anyone has a pointer to work (preferably in a qualitative reasoning system) that discusses the use of multi-layer constraint propogation? thanks much Jim Hendler Ass't Professor U of Md College Park, Md. 20742 [I would check into the MIT work (don't have the reference handy, but some of it's in the two-volume AI: An MIT Perspective) on modeling electronic circuits. All but the first papers used multiple views of subsystems to permit propagation of constraints at different granularities. Subsequent work on electronic fault diagnosis (e.g., Randy Davis) goes even further. Other work in "pyramidal" parsing (speech, images, line drawings) has grown from the Hearsay blackboard architecture. -- KIL] ------------------------------ Date: 27 Oct 86 14:13 EST From: SHAFFER%SCOVCB.decnet@ge-crd.arpa Subject: Model-base Reasoning Hello: I am looking for articles and books which describe the theory of Model-based Reasoning, MBR. Here at GE we have an interest in MBR for our next generation of KEE-based ESEs. I will publish a summary of my findings sometime in the future. Also, I would be interested in any related topics which related to MBR and its uses. Thanks, Earl Shaffer GE - VFSC - Bld 100 Po Box 8555 Phila, PA 19101 ------------------------------ Date: 28 Oct 86 09:52 EST From: SHAFFER%SCOVCB.decnet@ge-crd.arpa Subject: Monotonic Reasoning I am somewhat new to AI and I am confused about the definition of "non-monotonic" reasoning, as in the documentation in Inference's ART system. It say that features allow for non-monotonic reasoning, but does not say what that type of reasoning is, or how it differs from monotonic reasoning, if there is such a thing. Earl Shaffer GE - VFSC Po box 8555 Phila , Pa 19101 [Monotonic reasoning is a process of logical inference using only true axioms or statements. Nonmonotonic reasoning uses statements believed to be true, but which may later prove to be false. It is therefore necessary to keep track of all chains of support for each conclusion so that the conclusion can be revoked if its basis statements are revoked. Other names for nonmonotonic reasoning are default reasoning and truth maintenance. -- KIL] ------------------------------ Date: 28 Oct 86 21:05:49 GMT From: uwslh!lishka@rsch.wisc.edu (a) Subject: Re: simulating a neural network I just read an interesting short blurb in the most recent BYTE issue (the one with the graphics board on the cover)...it was in Bytelines or something. Now, since I skimmed it, my info is probably a little sketchy, but here's about what it said: Apparently Bell Labs (I think) has been experimenting with neural network-like chips, with resistors replacing bytes (I guess). They started out with about 22 'neurons' and have gotten up to 256 or 512 (can't remember which) 'neurons' on one chip now. Apparently these 'neurons' are supposed to run much faster than human neurons...it'll be interesting to see how all this works out in the end. I figured that anyone interested in the neural network program might be interested in the article...check Byte for actual info. Also, if anyone knows more about this experiment, I would be interested, so please mail me any information at the below address. -- Chris Lishka /l lishka@uwslh.uucp Wisconsin State Lab of Hygiene -lishka%uwslh.uucp@rsch.wisc.edu \{seismo, harvard,topaz,...}!uwvax!uwslh!lishka ------------------------------ Date: 27 Oct 86 19:50:58 GMT From: yippee.dec.com!glantz@decwrl.dec.com Subject: Re: Simulating neural networks ********************* Another good reference is: Martin, R., Lukton, A. and Salthe, S.N., "Simulation of Cognitive Maps, Concept Hierarchies, Learning by Simile, and Similarity Assessment in Homogeneous Neural Nets," Proceedings of the 1984 Summer Computer Simulation Conference, Society for Computer Simulation, vol. 2, 808-821. In this paper, Martin discusses (among other things) simulating the effects of neurotransmittors and inhibitors, which can have the result of generating goal-seeking behavior, which is closely linked to the ability to learn. Mike Glantz Digital Equipment Centre Technique Europe BP 29 Sophia Antipolis 06561 Valbonne CEDEX France My employer is not aware of this message. ********************* ------------------------------ Date: 27 Oct 86 17:36:23 GMT From: zeus!berke@locus.ucla.edu (Peter Berke) Subject: Glib "computation" In article <1249@megaron.UUCP> wendt@megaron.UUCP writes: >Anyone interested in neural modelling should know about the Parallel >Distributed Processing pair of books from MIT Press. They're >expensive (around $60 for the pair) but very good and quite recent. > >A quote: > >Relaxation is the dominant mode of computation. Although there >is no specific piece of neuroscience which compels the view that >brain-style computation involves relaxation, all of the features >we have just discussed have led us to believe that the primary >mode of computation in the brain is best understood as a kind of >relaxation system in which the computation proceeds by iteratively >seeking to satisfy a large number of weak constraints. Thus, >rather than playing the role of wires in an electric circuit, we >see the connections as representing constraints on the co-occurrence >of pairs of units. The system should be thought of more as "settling >into a solution" than "calculating a solution". Again, this is an >important perspective change which comes out of an interaction of >our understanding of how the brain must work and what kinds of processes >seem to be required to account for desired behavior. > >(Rumelhart & Mcclelland, Chapter 4) > Isn't 'computation' a technical term? Do R&Mc prove that PDP is equivalent to computation? Would Turing agree that "settling into a solution" is computation? Some people have tried to show that symbols and symbol processing can be represented in neural nets, but I don't think anyone has proved anything about the problems they purportedly "solve," at least not to the extent that Turing did for computers in 1936, or Church in the same year for lambda calculus. Or are R&Mc using 'computing' to mean 'any sort of machination whatever'? And is that a good idea? Church's Thesis, that computing and lambda-conversion (or whatever he calls it) are both equivalent to what we might naturally consider calcuable could be extended to say that neural nets "settle" into the same solutions for the same class of problems. Or, one could maintain, as neural netters tend to implicitly, that "settling" into solutions IS what we might naturally consider calculable, rather than being merely equivalent to it. These are different options. The first adds "neural nets" to the class of formalisms which can express solutions equivalent to each other in "power," and is thus a variant on Church's thesis. The second actually refutes Church's Thesis, by saying this "settling" process is clearly defined and that it can realize a different (or non-comparable) class of problems, in which case computation would not be (provably) equivalent to it. Of course, if we could show BOTH that: (1) "settling" is equivalent to "computing" as formally defined by Turing, and (2) that "settling" IS how brains work, then we'd have a PROOF of Church's Thesis. Until that point it seems a bit misleading or misled to refer to "settling" as "computation." Peter Berke ------------------------------ End of AIList Digest ******************** From vtcs1::in% Fri Oct 31 02:05:43 1986 Date: Fri, 31 Oct 86 02:05:35 est From: vtcs1::in% (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #241 Status: R AIList Digest Thursday, 30 Oct 1986 Volume 4 : Issue 241 Today's Topics: Philosophy & Physics - Analog/Digital Distinction ---------------------------------------------------------------------- Date: 27 Oct 86 06:08:33 GMT From: rutgers!princeton!mind!harnad@titan.arc.nasa.gov Subject: Re: Defining the Analog/Digital Distinction Tom Dietterich (orstcs!tgd) responds as follows to my challenge to define the A/D distinction: > In any representation, certain properties of the representational > medium are exploited to carry information. Digital representations > tend to exploit fewer properties of the medium. For example, in > digital electronics, a 0 could be defined as anything below .2 volts > and a 1 as anything above 4 volts. This is a simple distinction. > An analog representation of a signal (e.g., in an audio amplifier) > requires a much finer grain of distinctions--it exploits the > continuity of voltage to represent, for example, the loudness > of a sound. So far so good. Analog representations "exploit" more of the properties (e.g., continuity) of the "representational" (physical?) medium to carry information. But then is the difference between an A and a D representation just that one is more (exploitative) and the other less? Is it not rather that they carry information and/or represent in a DIFFERENT WAY? In what does that difference consist? (And what does "exploit" mean? Exploit for whom?) > A related notion of digital and analog can be obtained by considering > what kinds of transformations can be applied without losing > information. Digital signals can generally be transformed in more > ways--precisely because they do not exploit as many properties of the > representational medium. Hence, if we add .1 volts to a digital 0 as > defined above, the result will either still be 0 or else be undefined > (and hence [un]detectable). A digital 1 remains unchanged under > addition of .1 volts. However, the analog signal would be > changed under ANY addition of voltage. "Preserving information under transformations" also sounds like a good candidate. But it seems to me that preservation-under-transformation is (or ought to be) a two-way street. Digital representations may be robust within their respective discrete boundaries, but it hardly sounds information-preserving to lose all the information between .2 volts and 4 volts. I would think that the invertibility of analog transformations might be a better instance of information preservation than the irretrievable losses of A/D. And this still seems to side-step the question of WHAT information is preserved, and in what way, by analog and digital representations, respectively. And should we be focusing on representations in this discussion, or on transformations (A/A, A/D, D/D, D/A)? Finally, what is the relation between a digital representation and a symbolic representation? Please keep those definitions coming. Stevan Harnad {allegra, bellcore, seismo, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ Date: 27 Oct 86 11:53:06 est From: Walter Hamscher Subject: The Analog/Digital Distinction Date: 23 Oct 86 17:20:00 GMT From: hp-pcd!orstcs!tgd@hplabs.hp.com (tgd) Here is a rough try at defining the analog vs. digital distinction. [ * * * ] I don't read all the messages on AiList, so I may have missed something here: but isn't ``analog vs digital'' the same thing as ``continuous vs discrete''? Continuous vs discrete, in turn, can be defined in terms of infinite vs finite partitionability. It's a property of the measuring system, not a property of the thing being measured. ------------------------------ Date: 27 Oct 86 15:29:01 GMT From: alice!jj@ucbvax.Berkeley.EDU Subject: Re: Defining the Analog/Digital Distinction > From allegra!princeton!mind!harnad Wed Dec 31 19:00:00 1969 > > > Tom Dietterich (orstcs!tgd) responds as follows to my challenge to > define the A/D distinction: > > > In any representation, certain properties of the representational > > ... > > of a sound. > > So far so good. Analog representations "exploit" more of the properties > ... > for whom?) > > > A related notion of digital and analog can be obtained by considering > > ... > > changed under ANY addition of voltage. > > "Preserving information under transformations" also sounds like a good > ... > representation and a symbolic representation? > > Please keep those definitions coming. > > Stevan Harnad What a pleasant little bit of sophistry. Mr. Harnad asks for a defination of "digital" and "analog", both words used in a precise way in a particular literature. He also asks that we do not use other words used in that literature to write the defination. In other words, we are asked to define something precisely, in a languange that does not have precise values. I suggest the first chapter of Rabiner and Gold, all of Wozencraft and Jacobs, and perhaps a good general text on signal processing for starters. That will define the language. Then the defination can be made. Philosophy is wonderful, it doesn't have to have anything to do with reality. -- WOBEGON WON'T BE GONE, TEDDY BEAR PICNIC AT 11. "If you love that Politician, use your Loo, use your Loo" (ihnp4;allegra;research)!alice!jj ------------------------------ Date: 27 Oct 86 22:06:14 GMT From: husc6!Diamond!aweinste@eddie.mit.edu (Anders Weinstein) Subject: Re: The Analog/Digital Distinction: Soliciting Definitions Philosopher Nelson Goodman has distinguishes analog from digital symbol systems in his book _Languages_of_Art_. The context is a technical investigation into the peculiar features of _notational_ systems in the arts; that is, systems like musical notation which are used to DEFINE a work of art by dividing the instances from the non-instances. The following excerpts contain the relevant definitions: (Warning--I've left out a lot of explanatory text and examples for brevity) The second requirement upon a notational scheme, then, is that the characters be _finitely_differentiated_, or _articulate_. It runs: For every two characters K and K' and every mark m that does not belong to both, determination that m does not belong to K or that m does not belong to K' is theoretically possible. ... A scheme is syntactically dense if it provides for infinitely many characters so ordered that between each two there is a third. ... When no insertion of other characters will thus destroy density, a scheme has no gaps and may be called _dense_throughout_. In what follows, "throughout" is often dropped as understood... [in footnote:] I shall call a scheme that contains no dense subscheme "completely discontinuous" or "discontinuous throughout". ... The final requirement [including others not quoted here] for a notational system is semantic finite differentiation; that is for every two characters K and K' such that their compliance classes are not identical and every object h that does not comply with both, determination that h does not comply with K or that h does not comply with K' must be theoretically possible. [defines 'semantically dense throughout' and 'semantically discontinuous' to parallel the syntactic definitions]. And his analog/digital distinction: A symbol _scheme_ is analog if syntactically dense; a _system_ is analog if syntactically and semantically dense. ... A digital scheme, in contrast, is discontinuous throughout; and in a digital system the characters of such a scheme are one-one correlated with compliance-classes of a similarly discontinous set. But discontinuity, though implied by, does not imply differentiation...To be digital, a system must be not merely discontinuous but _differentiated_ throughout, syntactically and semantically... If only thoroughly dense systems are analog, and only thoroughly differentiated ones are digital, many systems are of neither type. To summarize: when a dense language is used to represent a dense domain, the system is analog; when a discrete (Goodman's "discontinuous") and articulate language maps a discrete and articulate domain, the system is digital. Note that not all discrete languages are "articulate" in Goodman's sense: Consider a language with only two characters, one of which contains all straight marks not longer than one inch and the other of which contains all longer marks. This is discrete but not articulate, since no matter how precise our tests become, there will always be a mark (infinitely many, in fact) that cannot be judged to belong to one or the other character. For more explanation, consult the source directly (and not me). Anders Weinstein ------------------------------ Date: 28 Oct 86 04:20:07 GMT From: allegra!princeton!mind!harnad@ucbvax.Berkeley.EDU Subject: The Analog/Digital Distinction Steven R. Jacobs (utah-cs!jacobs) of the University of Utah CS Dept has given me permission to post his contribution to defining the A/D distinction. It appears below, followed at the very end by some comments from me. [Will someone with access please post a copy to sci.electronics?] >> One prima facie non-starter: "continuous" vs. "discrete" physical processes. >I apologize if this was meant to avoid discussion of continuous/discrete >issues relating to analog/digital representations. I find it difficult >to avoid talking in terms of "continuous" and "discrete" processes when >discussing the difference between analog and digital signals. I am >approaching the question from a signal processing point of view, so I >tend to assume that "real" signals are analog signals, and other methods >of representing signals are used as approximations of analog signals (but >see below about a physicist's perspective). Yes, I realize you asked for >objective definitions. For my own non-objective convenience, I will use >analog signals as a starting point for obtaining other types of signals. >This will assist in discussing the operations used to derive non-analog >signals from analog signals, and in discussing the effects of the operations >on the mathematics involved when manipulating the various types of signals >in the time and frequency domains. > >The distinction of continuous/discrete can be applied to both the amplitude >and time axes of a signal, which allows four types of signals to be defined. >So, some "loose" definitions: > >Analog signal -- one that is continuous both in time and amplitude, so that > the amplitude of the signal may change to any amplitude at any time. > This is what many electrical engineers might describe as a "signal". > >Sampled signal -- continuous in amplitude, discrete in time (usually with > eqully-spaced sampling intervals). Signal may take on any amplitude, > but the amplitude changes only at discrete times. Sampled signals > are obtained (obviously?) by sampling analog signals. If sampling is > done improperly, aliasing will occur, causing a loss of information. > Some (most?) analog signals cannot be accurately represented by a > sampled signal, since only band-limited signals can be sampled without > aliasing. Sampled signals are the basis of Digital Signal Processing, > although digital signals are invariably used as an approximation of > the sampled signals. > >Quantized signal -- piece-wise continuous in time, discrete in amplitude. > Amplitude may change at any time, but only to discrete levels. All > changes in amplitude are steps. > >Digital signal -- one that is discrete both in time and amplitude, and may > change in (discrete) amplitude only at certain (discrete, usually > uniformly spaced) time intervals. This is obtained by quantizing > a sampled signal. > >Other types of signals can be made by combining these "basic" types, but >that topic is more appropriate for net.bizarre than for sci.electronics. > >The real distinction (in my mind) between these representations is the effect >the representation has on the mathematics required to manipulate the signals. > >Although most engineers and computer scientists would think of analog signals >as the most "correct" representations of signals, a physicist might argue that >the "quantum signal" is the only signal which corresponds to the real world, >and that analog signals are merely a convenient approximation used by >mathematicians. > >One major distinction (from a mathematical point of view) between sampled >signals and analog signals can be best visualized in the frequency domain. >A band-limited analog signal has a Fourier transform which is finite. A >sampled representation of the same signal will be periodic in the Fourier >domain. Increasing the sampling frequency will "spread out" the identical >"clumps" in the FT (fourier transform) of a sampled signal, but the FT >of the sampled signal will ALWAYS remain periodic, so that in the limit as >the sampling frequency approaches infinity, the sampled signal DOES NOT >become a "better" approximation of the analog signal, they remain entirely >distinct. Whenever the sampling frequency exceeds the Nyquist frequency, >the original analog signal can be exactly recovered from the sampled signal, >so that the two representations contain the equivalent information, but the >two signals are not the same, and the sampled signal does not "approach" >the analog signal as the sampling frequency is increased. For signals which >are not band-limited, sampling causes a loss of information due to aliasing. >As the sampling frequency is increased, less information is lost, so that the >"goodness" of the approximation improves as the sampling frequency increases. >Still, the sampled signal is fundamentally different from the analog signal. >This fundamental difference applies also to digital signals, which are both >quantized and sampled. > >Digital signals are usually used as an approximation to "sampled" signals. >The mathematics used for digital signal processing is actually only correct >when applied to sampled signals (maybe it should be called "Sampled Signal >Processing" (SSP) instead). The approximation is usually handled mostly by >ignoring the "quantization noise" which is introduced when converting a >sampled analog signal into a digital signal. This is convenient because it >avoids some messy "details" in the mathematics. To properly deal with >quantized signals requires giving up some "nice" properties of signals and >operators that are applied to signals. Mostly, operators which are applied >to signals become non-commutative when the signals are discrete in amplitude. >This is very much related to the "Heisenburg uncertainty principle" of >quantum mechanics, and to me represents another "true" distinction between >analog and digital signals. The quantization of signals represents a loss of >information that is qualitatively different from any loss of information that >occurs from sampling. This difference is usally glossed over or ignored in >discussions of signal processing. > >Well, those are some half-baked ideas that come to my mind. They are probably >not what you are looking for, so feel free to post them to /dev/null. > >Steve Jacobs > - - - - - - - - - - - - - - - - - - - - - - - - REPLY: > I apologize if this was meant to avoid discussion of continuous/discrete > issues relating to analog/digital representations. It wasn't meant to avoid discussion of continuous/discrete at all; just to avoid a simple-minded equation of C/D with A/D, overlooking all the attendant problems of that move. You certainly haven't done that in your thoughtful and articulate review and analysis. > I tend to assume that "real" signals are analog signals, and other > methods of representing signals are used as approximations of analog > signals. That seems like the correct assumption. But if we shift for a moment from considering the A or D signals themselves and consider instead the transformation that generated them, the question arises: If "real" signals are analog signals, then what are they analogs of? Let's borrow some formal jargon and say that there are (real) "objects," and then there are "images" of them under various types of transformations. One such transformation is an analog transformation. In that case the image of the object under the (analog) transformation can also be called an "analog" of the object. Is that an analog signal? The approximation criterion also seems right on the mark. Using the object/transformation/image terminology again, another kind of a transformation is a "digital" transformation. The image of an object (or of the analog image of an object) under a digital transformation is "approximate" rather than "exact." What is the difference between "approximate" and "exact"? Here I would like to interject a tentative candidate criterion of my own: I think it may have something to do with invertibility. A transformation from object to image is analog if (or to the degree that) it is invertible. In a digital approximation, some information or structure is irretrievably lost (the transformation is not 1:1). So, might invertibility/noninvertibility have something to do with the distinction between an A and a D transformation? And do "images" of these two kinds count as "representations" in the sense in which that concept is used in AI, cognitive psychology and philosophy (not necessarily univocally)? And, finally, where do "symbolic" representations come in? If we take a continuous object and make a discrete, approximate image of it, how do we get from that to a symbolic representation? > Analog signal -- one that is continuous both in time and amplitude. > Sampled signal -- continuous in amplitude, discrete in time... > If sampling is done improperly, aliasing will occur, causing a > loss of information. > Quantized signal -- piece-wise continuous in time, discrete in > amplitude. > Digital signal -- one that is discrete both in time and amplitude... > This is obtained by quantizing a sampled signal. Both directions of departure from the analog, it seems, lose information, unless the interpolations of the gaps in either time or amplitude can be accurately made somehow. Question: What if the original "object" is discrete in the first place, both in space and time? Does that make a digital transformation of it "analog"? I realize that this is violating the "signal" terminology, but, after all, signals have their origins too. Preservation and invertibility of information or structure seem to be even more general features than continuity/discreteness. Or perhaps we should be focusing on the continuity/noncontinuity of the transformations rather than the objects? > a physicist might argue that the "quantum signal" is the only > signal which corresponds to the real world, and that analog > signals are merely a convenient approximation used by mathematicians. This, of course, turns the continuous/discrete and the exact/approximate criteria completely on their heads, as I think you recognize too. And it's one of the things that makes continuity a less straightforward basis for the A/D distinction. > Mostly, operators which are applied to signals become > non-commutative when the signals are discrete in amplitude. > This is very much related to the "Heisenburg uncertainty principle" > of quantum mechanics, and to me represents another "true" distinction > between analog and digital signals. The quantization of signals > represents a loss of information that is qualitatively different from > any loss of information that occurs from sampling. I'm not qualified to judge whether this is an anolgy or a true quantum effect. If the latter, then of course the qualitative difference resides in the fact that (on current theory) the information is irretrievable in principle rather than merely in practice. > Well, those are some half-baked ideas that come to my mind. Many thanks for your thoughtful contribution. I hope the discussion will continue "baking." Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ Date: 27 Oct 86 03:29:00 GMT From: uiucuxe!goldfain@uxc.cso.uiuc.edu Subject: Re: The Analog/Digital Distinction: Sol Analog devices/processes are best viewed as having a continuous possible range of values. (An interval of the real line, for example.) Digital devices/processes are best viewed as having an underlying granularity of discrete possible values. (Representable by a subset of the integers.) ----------------- This is a pretty good definition, whether you like it or not. I am curious as to what kind of discussion you are hoping to get, when you rule out the correct distinction at the outset ... ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Tue Nov 4 02:14:22 1986 Date: Tue, 4 Nov 86 02:14:02 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #242 Status: R AIList Digest Thursday, 30 Oct 1986 Volume 4 : Issue 242 Today's Topics: Philosophy - Searle, Turing, Symbols, Categories ---------------------------------------------------------------------- Date: 27 Oct 86 03:58:54 GMT From: spar!freeman@decwrl.dec.com Subject: Re: Searle, Turing, Symbols, Categories In article <12@mind.UUCP> harnad@mind.UUCP (Stevan Harnad) writes: > >freeman@spar.UUCP (Jay Freeman) replies: > >> Possibly a more interesting test [than the robotic version of >> the Total Turing Test] would be to give the computer >> direct control of the video bit map and let it synthesize an >> image of a human being. > > Manipulating digital "images" is still only symbol-manipulation. [...] Very well, let's equip the robot with an active RF emitter so it can jam the camera's electronics and impose whatever bit map it wishes, whether the camera likes it or not. Too silly? Very well, let's design a robot in the shape of a back projector, and let it create internally whatever representation of a human being it wishes the camera to see, and project it on its screen for the camera to pick up. Such a robot might do a tolerable job of interacting with other parts of the "objective" world, using robot arms and whatnot of more conventional design, so long as it kept them out of the way of the camera. Still too silly? Very well, let's create a vaguely anthropomorphic robot and equip its external surfaces with a complete covering of smaller video displays, so that it can achieve the minor details of human appearance by projection rather than by mechanical motion. (We can use a crude electronic jammer to limit the amount of detail that the camera can see, if necessary.) Well, maybe our model shop is good enough to do most of the details of the robot convincingly, so we'll only have to project subtle details of facial expression. Maybe just the eyes. Slightly more seriously, if you are going to admit the presence of electronic or mechanical devices between the subject under test and the human to be fooled, you must accept the possibility that the test subject will be smart enough to detect their presence and exploit their weaknesses. Returning to a more facetious tone, consider a robot that looks no more anthropomorphic than your vacuum cleaner, but that is possessed of moderate manipulative abilities and a good visual perceptive apparatus, and furthermore, has a Swiss Army knife. Before the test commences, the robot sneakily rolls up to the camera and removes the cover. It locates the connections for the external video output, and splices in a substitute connection to an external video source which it generates. Then it replaces the camera cover, so that everything looks normal. And a test time, the robot provides whatever image it wants the testers to see. A dumb robot might have no choice but to look like a human being in order to pass the test. Why should a smart one be so constrained? -- Jay Freeman ------------------------------ Date: Mon 27 Oct 86 20:02:39-EST From: Albert Boulanger Subject: Turing Test I think it is amusing and instructive to look at real attempts of the turing test. One interesting attempt is written up in the post scriptum of the chapter: "A Coffeehouse Conversation on the Turing Test" Metamagical Themas Douglas Hofstadter Basic Books 1985 Albert Boulanger BBN Labs ------------------------------ Date: 27 Oct 86 17:23:31 GMT From: rutgers!princeton!mind!harnad@titan.arc.nasa.gov Subject: Pseudomath about the Turing Test: Reply to Padin [Until the problem of follow-up articles to mod.ai through Usenet is straightened out, I'm temporarily responding to mod.ai on net.ai.] In mod.ai, in Message-ID: <8610270723.AA05463@ucbvax.Berkeley.EDU>, under the subject heading THE PSEUDOMATH OF THE TURING TEST, PADIN@FNALB.BITNET writes: > DEFINE THE SET Q={question1,question2,...}. LETS NOTE THAT > FOR EACH q IN Q, THERE IS AN INFINITE NUMBER OF RESPONSES (THE > RESPONSES NEED NOT BE RELEVANT TO THE QUESTION, THEY JUST NEED TO BE > RESPONSES). IN FACT, WE CAN DEFINE A SET R={EVERY POSSIBLE RESPONSE TO > ANY QUESTION}, i.e., R={r1,r2,r3,...}. Do pseudomath and you're likely to generate pseudoproblems. Nevertheless, this way of formulating it does inadvertently illustrate quite clearly why the symbolic version of the turing test is inadequate and the robotic version is to be preferred. The symbolic version is equivalent to the proverbial monkey's chances of typing Shakespeare by combinatorics. The robotic version (pending the last word on basic continuity/discontinuity in microphysics) is then no more or less of a combinatorial problem than Newtonian Mechanics. [Concerning continuity/discreteness, join the ongoing discussion on the A/D distinction that's just started up in net/mod.ai.] > THE EXISTENCE OF ...A FUNCTION T THAT MAPS A QUESTION q TO A SET > OF RESPONSES RR... FOR ALL QUESTIONS q IS EVIDENCE FOR THE PRESENCE > OF MIND SINCE T CHOOSES, OUT OF AN INFINITE NUMBER OF RESPONSES, > THOSE RESPONSES THAT ARE APPROPRIATE TO AN ENTITY WITH A MIND. Pare off the pseudomath about "choosing among infinities" and you just get a restatement of the basic intuition behind the turing test: That an entity has a mind if it acts indistinguishably from an entity with a mind. > NOW A PROBLEM [arises]: WHO IS TO DECIDE WHICH SUBSET OF RESPONSES > INDICATES THE EXISTENCE OF MIND? WHO WILL DECIDE WHICH SET IS > APPROPRIATE TO INDICATE AN ENTITY OTHER THAN OURSELVES IS OUT THERE > RESPONDING? The same one who decides in ongoing, everyday "solutions" to the other-minds problem. And on exactly the same basis: indistinguishability of performance. > [If] WE GET A RESPONSE WHICH APPEARS TO BE RANDOM, IT WOULD SEEM THAT > THIS WOULD BE SUFFICIENT TO LABEL [the] RESPONDENT A MINDLESS ENTITY. > HOWEVER, IT IS THE EXACT RESPONSE ONE WOULD EXPECT OF A SCHIZOPHRENIC. When will this tired prima facie objection (about schizophrenia, retardation, aphasia, coma, etc.) at last be laid to rest? Damaged humans inherit the benefit of the doubt from what we know about their biological origins AND about the success of their normal counterparts in passing the turing test. Moreover, there is no problem in principle with subhuman or nonhuman performance -- in practice we turing-test animals too -- and this too is probably parasitic on our intuitions about normal human beings (although the evolutionary order was probably vice versa). Also, schizophrenics don't just behave randomly; if a candidate just behaved randomly it would not only justifiably flunk the turing test, but it would not survive either. (I don't even know what behaving purely randomly might mean; it seems to me the molecules would never make it through embryogeny...) On the other hand, which of us doesn't occasionally behave randomly, and some more often than other?. We can hardly expect the turing test to provide us with the criteria for extreme conditions such as brain death if even biologists have problems with that. All these exotic variants are pseudoproblems and red herrings, especially when we are nowhere in our progress in developing a system that can give the normal version of the turing test a run for its money. > NOW IF WE ARE TO USE OUR JUDGEMENT IN DETERMINING THE PRESENCE OF > ANOTHER MIND, THEN WE MUST ACCEPT THE POSSIBILITY OF ERROR INHERENT > IN THE HUMAN DECISION MAKING PROCESS. AT BEST,THEN, THE TURING TEST > WILL BE ABLE TO GIVE US ONLY A HINT AT THE PRESENCE OF ANOTHER MIND; > A LEVEL OF PROBABILITY. What else is new? Even the theories of theoretical physics are only true with high probability. There is no mathematical proof that our inferences are entailed with necessity by the data. This is called "underdetermination" and "inductive risk," and it is endemic to all empirical inquiry. But besides that, the turing test has even a second layer of underdermination that verges on indeterminacy. I have argued that it has two components: One is the formal theorist's task of developing a device that can generate all of our performance capacities, i.e.,one that can pass the Total Turing Test. So far, with only "performance capacity" having been mentioned, the level of underdetermination is that of ordinary science (it may have missed some future performance capacity, or it may fail tomorrow, or it may just happen to accomplish the same performance in a radically different way, just as the universe may happen to differ from our best physical theory). The second component of the turing test, however, is informal, intuitive and open-ended, and it's the one we usually have in mind when we speak of the turing test: Will a normal human being be able to tell the candidate apart from someone with a mind? The argument is that turing-indistinguishability of (total) performance is the only basis for making that judgment in any case. Fallible? Of course that kind of judgment is fallible. Certainly no less fallible than ordinary scientific inference; and (I argue) no more fallible than our judgments about other minds. What more can one ask? Apart from the necessary truths of mathematics, the only other candidate for a nonprobabilistic certainty is our direct ("incorrigible") awareness of our OWN minds (although even there the details seem a bit murky...). Stevan Harnad {allegra, bellcore, seismo, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ Date: 28 Oct 86 08:40:00 EDT From: "CUGINI, JOHN" Reply-to: "CUGINI, JOHN" Subject: son of yet more wrangling on Searle, Turing, Quine, Hume, ... Warning: the following message is long and exceeds the FDA maximum daily recommended dosage of philosophizing. You have been warned. This is the exchange that kicked off this whole diatribe: >>> Harnad: there is no rational reason for being more sceptical about robots' >>> minds (if we can't tell their performance apart from that of people) >>> than about (other) peoples' minds. >> Cugini: One (rationally) believes other people are conscious BOTH because >> of their performance and because their internal stuff is a lot like >> one's own. > This is a very important point and a subtle one, so I want to make > sure that my position is explicit and clear: I am not denying that > there exist some objective data that correlate with having a mind > (consciousness) over and above performance data. In particular, > there's (1) the way we look and (2) the fact that we have brains. What > I am denying is that this is relevant to our intuitions about who has a > mind and why. I claim that our intuitive sense of who has a mind is > COMPLETELY based on performance, and our reason can do no better. These > other correlates are only inessential afterthoughts, and it's irrational > to take them as criteria. This riposte seems implausible on the face of it. You seem to want to pretend that we know absolutely nothing about the basis of thought in humans, and to "suppress" all evidence based on such knowledge. But that's just wrong. Brains *are* evidence for mind, in light of our present knowledge. > My supporting argument is very simple: We have absolutely no intuitive > FUNCTIONAL ideas about how our brains work. (If we did, we'd have long > since spun an implementable brain theory from our introspective > armchairs.) Consequently, our belief that brains are evidence of minds and > that the absence of a brain is evidence of the absence of a mind is based > on a superficial black-box correlation. It is no more rational than > being biased by any other aspect of appearance, such as the color of > the skin, the shape of the eyes or even the presence or absence of a tail. Hoo hah, you mean to say that belief based on "black-box correlation" is irrational in the absence of a fully-supporting theoretical framework? Balderdash. People in, say, 1500 AD were perfectly rational in predicting tides based on the position of the moon (and vice-versa) even though they hadn't a clue as to the mechanism of interaction. If you keep asking "why" long enough, *all* science is grounded on such brute-fact correlation (why do like charges repel, etc.) - as Hume pointed out a while back. > To put it in the starkest terms possible: We wouldn't know what device > was and was not relevantly brain-like if it was staring us in the face > -- EXCEPT IF IT HAD OUR PERFORMANCE CAPACITIES (i.e., it could pass > the Total Turing Test). That's the only thing our intuitions have to > go on, and our reason has nothing more to offer either. Except in the case of actual other brains (which are, by definition, relevantly brain-like). The only skepticism open to one is that one's own brain is unique in its causal powers - possible, but hardly the best rational hypothesis. > People were sure (as sure as they'll ever be) that other people had > minds long before they ever discovered they had brains. I myself believed > the brain was just a figure of speech for the first dozen or so years of > my life. Perhaps there are people who don't learn or believe the news > throughout their entire lifetimes. Do you think these people KNOW any > less than we do about what does or doesn't have a mind? ... Let me re-cast Harnad's argument (perhaps in a form unacceptable to him): We can never know any mind directly, other than our own, if we take the concept of mind to be something like "conscious intelligence" - ie the intuitive (and correct, I believe) concept, rather than some operational definition, which has been deliberately formulated to circumvent the epistemological problems. (Harnad, to his credit, does not stoop to such positivist ploys.) But the only external evidence we are ever likely to get for "conscious intelligence" is some kind of performance. Moreover, the physical basis for such performance will be known only contingently, ie we do not know, a priori, that it is brains, rather than automatic dishwashers, which generate mind, but rather only as an a posteriori correlation. Therefore, in the search for mind, we should rely on the primary criterion (performance), rather than on such derivative criteria as brains. I pretty much agree with the above account except for the last sentence which prohibits us from making use of derivative criteria. Why should we limit ourselves so? Since when is that part of rationality? No, the fact is we do have more reason to suppose mind of other humans than of robots, in virtue of an admittedly derivative (but massively confirmed) criterion. And we are, in this regard, in a epistemological position *superior* to those who don't/didn't know about such things as the role of the brain, ie we have *more* reason to believe in the mindedness of others than they do. That's why primitive tribes (I guess) make the *mistake* of attributing mind to trees, weather, etc. Since raw performance is all they have to go on, seemingly meaningful activity on the part of any old thing can be taken as evidence of consciousness. But we sophisticates have indeed learned a thing or two, in particular, that brains support consciousness, and therefore we (rationally) give the benefit of the doubt to any brained entity, and the anti-benefit to un-brained entities. Again, not to say that we might not learn about other bases for mind - but that hardly disparages brainedness as a rational criterion for mindedness. Another point, which I'll just state rather than argue for is that even performance is only *contingently* a criterion for mind - ie, it so happens, in this universe, that mind often expresses itself by playing chess, etc., just as it so happens that brains cause minds. And so there's really not much difference between relying on one contingent correlate (performance) rather than another (brains) as evidence for the presence of mind. > > Why is consciousness a red herring just because it adds a level > > of uncertainty? > > Perhaps I should have said indeterminacy. If my arguments for > performance-indiscernibility (the turing test) as our only objective > basis for inferring mind are correct, then there is a level of > underdetermination here that is in no way comparable to that of, say, > the unobservable theoretical entities of physics (say, quarks, or, to > be more trendy, perhaps strings). Ordinary underdetermination goes > like this: How do I know that your theory's right about the existence > and presence of strings? Because WITH them the theory succeeds in > accounting for all the objective data (let's pretend), and without > them it does not. Strings are not "forced" by the data, and other > rival theories may be possible that work without them. But until these > rivals are put forward, normal science says strings are "real" (modulo > ordinary underdetermination). > Now try to run that through for consciousness: How do I know that your > theory's right about the existence and presence of consciousness (i.e., > that your model has a mind)? "Because its performance is > turing-indistinguishable from that of creatures that have minds." Is > your theory dualistic? Does it give consciousness an independent, > nonphysical, causal role? "Goodness, no!" Well then, wouldn't it fit > the objective data just as well (indeed, turing-indistinguishably) > without consciousness? "Well..." > That's indeterminacy, or radical underdetermination, or what have you. > And that's why consciousness is a methodological red herring. I admit, I have trouble following the line of argument above. Is this Quine's "it's real if it's a term in our best-confirmed theories" approach? But I think Quine is quite wrong, if that is his assertion. I know consciousness (my own, at least) exists, not as some derived theoretical construct which explains low-level data (like magnetism explains pointer readings), but as the absolutely lowest rock-bottom datum there is. Consciousness is the data, not the theory - it is the explicandum, not the explicans (hope I got that right). It's true that I can't directly observe the consciousness of others, but so what? That's an epistemological inconvenience, but it doesn't make consciousness a red herring. > I don't know what you mean, by the way, about always being able to > "engineer anything with anything at some level of abstraction." Can > anyone engineer something to pass the robotic version of the Total > Turing Test right now? And what's that "level of abstraction" stuff? > Robots have to do their thing in the real world. And if my > groundedness arguments are valid, that ain't all done with symbols > (plus add-on peripheral modules). The engineering remark was to re-inforce the idea that, perhaps, being-composed-of-protein might not be as practically incidental as many assume. Frinstance, at some level of difficulty, one can get energy from sunlight "as plants do." But the issues are: do we get energy from sunlight in the same way? How similar do we demand that the processes are? It might be easy to be as efficient as plants in getting energy from sunlight through non-biological technology. But if we're interested in simulation at a lower level of abstraction, eg, photosynthesis, then, maybe, a non-biological approach will be impractical. The point is we know we can simulate human chess-playing abilities with non-biological technology. Should we just therefore declare the battle for mind won, and go home? Or ask the harder question: what would it take to get a machine to play a game of chess like a person does, ie, consciously. BTW, I quite agree with your more general thesis on the likely inadequacy of symbols (alone) to capture mind. John Cugini ------------------------------ End of AIList Digest ******************** ------- From in%@vtcs1 Tue Nov 4 02:14:36 1986 Date: Tue, 4 Nov 86 02:14:17 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #243 Status: R AIList Digest Monday, 3 Nov 1986 Volume 4 : Issue 243 Today's Topics: Games - Chess, Seminars - Aid To Database Design (UPenn) & A Circumscriptive Theory of Plan Recognition (BBN), Conferences - Sydney Expert Systems Conference & European Conference on Object Oriented Programming ---------------------------------------------------------------------- Date: 28 Oct 86 11:35:39 GMT From: mcvax!unido!ab@seismo.css.gov (ab) Subject: Re: Places in Vancouver? Really Chess programs. > I don't really understand why there are not any really good chess > programs available for home computers. Fidelity has a machine > with an official USCF rating of 2100 for 200 bucks. I am pretty > sure that this has an 8 bit processor. Someone should be able to > come up with a 68k program that is better than this! Did you hear of the recent PSION-CHESS program for the Atari ST? This is a completely new program developed by Richard Lang. It uses heuristic search instead of the alpha-beta-procedure. This means that the program can examine the game tree to arbitrary depth. It uses a highly selective search to investigate the interesting lines of play. Moreover its playing style is very aggressive. The search concentrates on lines of play which are tactically sharp and which force the opponent to play in a way which can be easily predicted. So not necessarily the best move is played but the tactically sharpest with reasonable outcome. This means that a depth of up to 20 plies can be forced and a gain of material in let's say 8 plies is recognized. The program can display up to 8 plies of its current expected moves. There exist two ways of displaying the board: 3d and 2d. You can set the board to an arbitrary position and there exist levels of play from novice (1 sec) to expert (same time) and infinity. Also there are problem modes for forced check mates. The program normally 'thinks' while its opponent has to move, but with the feature 'handicap' this can be disabled. A lot of other features are supported which could be mentioned. It seems to me that this program is identical to the Mephisto Munchen with Amsterdam-modul since that one also uses the same strategy, the same processor and is also by Richard Lang. If true that would mean that PSION-CHESS alias Mephisto-Munchen is the recent world champion of microcomputer chess (championship in Amsterdam fall 1985). Has anyone further information on this program or on its strength? I am particularly interested in the new programing approach realized in this program. There exist some articles by Larry R. Harris about heuristic search in chess, but these articles date back to 1975. Are there other available programs which use the new approach? Andreas Bormann University of Dortmund [UniDo] West Germany Uucp: ab@unido.uucp Path: {USA}!seismo!{mcvax}!unido!ab {Europe}!{cernvax,diku,enea,ircam,mcvax,prlb2,tuvie,ukc}!unido!ab Bitnet: ab@unido.bitnet (== ab@ddoinf6.bitnet) [ Followups will be directed to net.games.chess only.] [ Any thoughts or opinions which may or may not have been expressed ] [ herein are my own. They are not necessarily those of my employer. ] [ Also I have no ambitions to sell PSION-CHESS or Mephisto computers.] ------------------------------ Date: Tue, 28 Oct 86 23:06 EST From: Tim Finin Subject: Seminar - Aid To Database Design (UPenn) Dissertation Defense Aid To Database Design: An Inductive Inference Approach Sitaram Lanka The conventional approach to the design of databases has the drawback that to specify a database schema, it requires the user interested in designing a schema to have the knowledge about both the domain and the data model. The aim of this research is to propose a semi automated system which designs a database schema in which the user need only have the knowledge of the underlying domain. This is expressed in terms of the information retrieval requirements that the database has to satisfy eventually. We have cast this as a problem in inductive inference where the input is in the form of Natural Language English queries. A database schema is inferred from this and is expressed in the functional data model. The synthesis of the database schema from the input queries is carried out by an inference mechanism. The central idea in designing the inference mechanism is the notion of compositionality and we have described it in terms of attribute grammars due to Kunth. A method has been proposed to detect any potentially false hypothesis that the inference mechanism may put forth and we have proposed a scheme to refine them such that we will obtain acceptable hypothesis. A prototype has been implemented on the Symbolics Lisp machine. Committee Dr. P. Buneman Dr. T. Finin (chairman) Dr. R. Gerritsen Supervisor Dr. A.K. Joshi Supervisor Dr. R.S. Nikhil Dr. B. Webber Date: October 31, 1986 Time: 2:30 pm Location: Room 23 ------------------------------ Date: Fri, 31 Oct 86 20:47:49 EST From: "Steven A. Swernofsky" Subject: Seminar - A Circumscriptive Theory of Plan Recognition (BBN) From: Brad Goodman BBN Laboratories Science Development Program AI/Education Seminar Speaker: Henry Kautz Dept. of Computer Science, University of Rochester (Henry@Rochester.Arpa) Title: A CIRCUMSCRIPTIVE THEORY OF PLAN RECOGNITION Date: 10:30a.m., Thursday, November 20th Location: 3rd floor large conference room, BBN Laboratories Inc., 10 Moulton St., Cambridge Abstract A plan library specifies the abstraction and decomposition relations between actions. A typical first-order representation of such a library does not, by itself, provide grounds for recognizing an agent's plans, given observations of the agent's actions. Several additional assumptions are needed: that the abstraction hierarchy is complete; that the decomposition hierarchy is complete; and that the agent's actions are, if possible, all part of the same plan. These assumptions are developed through the construction of a certain class of minimal models of the plan library. Circumscription provides a general non-constructive method for specifying a class of minimal models. For the specific case at hand, however, we can mechanically generate a set of first-order axioms which precisely capture the assumptions. The result is a "competence theory" of plan recognition, which correctly handles such difficult matters as disjunctive observations and multiple plans. The theory may be partially implemented by efficient (but limited) algorithms. ------------------------------ Date: Mon, 27 Oct 86 17:43:23 EST From: Jason Catlett Subject: Call for Papers, Sydney Expert Systems Conference >From moncskermit!munnari!seismo!ut-sally!pyramid!hplabs!hplabsc!taylor >From: taylor@hplabsc.UUCP (Dave Taylor) Newsgroups: mod.conferences Subject: Call-For-Papers: Sydney Expert Systems Conference Location: Sydney, Australia CALL FOR PAPERS The Third Australian Conference on Applications of Expert Systems Sydney, 13-15 May The Sydney Expert Systems Group has organised two successful conferences on this theme, including keynote addresses from internationally-recognised authorities Bruce Buchanan (Stanford University), Donald Michie (Turing Institute), Neil Pundit (Digital Equipment Corporation, USA), Donald Waterman (Rand Corporation) and Patrick Winston (M.I.T.). The 1987 conference will continue this tradition, with addresses from distinguished overseas speakers and Australian experts. Papers are invited on any aspect of expert systems technology, including - examples of expert systems that have been developed for particular applications - design and evaluation of tools for building expert systems - knowldege engineering methodology - specialised hardware for expert systems Contributions that discuss the authors' experiences/successes/ lessons learned in building expert systems will be particularly welcome. Papers of any size will be considered but a length of 15-30 pages is recommended. All accepted papers will be published in the Proceedings. Authors should note the following dates: Deadline for papers: 30th January 1987 Notification of acceptance: 13th March 1987 Deadline for camera-ready copy: 10th April 1987 Presentation of paper: 13-15th May 1987 Papers should be sent to the Program Chairman, Dr J. R. Quinlan School of Computing Sciences NSW Institute of Technology Broadway NSW 2007 Australia Requests for registration forms should be sent to "ES Conference Registrations, c/o Dr John Debenham" at the above address. ------------------------------ Date: Thu, 30 Oct 1986 10:50 EST From: HENRY%OZ.AI.MIT.EDU@XX.LCS.MIT.EDU Subject: European Conference on Object Oriented Programming Date: Mon, 6 Oct 86 18:25:23 -0100 (MET) From: pierre Cointe To: henry at ai.ai.mit.edu EUROPEAN CONFERENCE ON OBJECT ORIENTED PROGRAMMING Call for Papers Paris, France: June 15-17 1987 Following the AFCET group's three previous Working Sessions on Object Oriented Languages next encounter will take place at the Centre Georges Pompidou (Paris) on June 15th, 16th & 17th 1987. With regard to the success of the previous workshops and to the increasing interest on the subject, the next meeting will be an international conference organized by AFCET. The program committee is: G. Attardi, DELPHI, Italy J. Bezivin, LIB (UBO & ENSTbr), France P. Cointe, CMI & LITP, France S. Cook, London University, England J.M. Hullot, INRIA, France B. Kristensen, Aalborg University Center, Denmark H. Lieberman, MIT, USA L. Steels, Brussels University, Belgium H. Stoyan, Konstanz University, West German B. Stroustrup, AT&T Bell Labs, USA J. Vaucher, Montreal University, Canada A. Yonezawa, Tokyo Institut of Technology, Japan The conference will consist of a presentation of selected papers. Well-known researchers having made major contributions in the field - like C. Hewitt and K. Nygaard - will also give invited lectures. This new conference will deal with all domains using the techniques and methodologies of Object Oriented Programming. It is likely to interest both software designers and users. Proposed themes are the following: - Theory : semantic models (instantiation, inheritance), compilation - Conception : new languages, new hardwares, new extensions of languages - Applications : man/machine interfaces, simulation, knowledge representation, data bases, operating systems - Methodology : Smalltalk-80 methodology, actor methodology, frame methodology, the abstract type approach - Development : industrial applications. The papers must be submitted in English and should not be longer than ten pages. Five copies must be received at one of the address below, no later than January 9th,1987 (and, if possible, by electronic mails to the conference co-chairmen). Papers selection will be done by circulating papers to members of the program committee having appropriate expertise. Authors will be notified of acceptance by February, 15th 1987. To be included in the Proceedings the definitive version of the paper must reach the AFCET office before April, 27th 1987. - Conference Co-chairmen - J.M. Hullot (INRIA) mcvax!inria!hullot - J. Bezivin (LIB) mcvax!inria!geocub!bezivin - Program Co-chairmen - P. Cointe (LITP) mcvax!inria!cointe - H. Lieberman (MIT) mcvax!ai.ai.mit.edu!henry - USA Coordinator - B. Stroustrup (AT&T, Bell Labs) mcvax!research!snb!bs Murray Hill, Nj 07974 USA (201 582 7393) - Organization - Claire Van Hieu AFCET 156 Boulevard Pereire 75017 Paris, France (1) 47.66.24.19 Following the conference - and in the same place - Jerome Chailloux and Christian Queinnec will organize on June 18th and 19th a workshop about Lisp and its standardization. People interested in Tutorials, Workshops or Exhibitions may contact the AFCET organization. ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Tue Nov 4 02:14:02 1986 Date: Tue, 4 Nov 86 02:13:46 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #244 Status: R AIList Digest Monday, 3 Nov 1986 Volume 4 : Issue 244 Today's Topics: Query - AI in Rehabilitation Med, AI Tools - Guru & PD Parser for Simple English Sentences, Representations - Music, Logic - Monotonicity, Review - Weizenbaum Keynote Address at U of Waterloo ---------------------------------------------------------------------- Date: 30 Oct 86 23:15:35 EST From: Steve blumenfrucht Subject: AIM in Rehabilitation Med I am trying to find people doing artificial intelligence work in the medical specialty of Physical Medicine and Rehabilitation. I am especially interested in finding MDs doing this. Help/suggestions are appreciated. Reply to BLUMENFRUCHT@RUTGERS ------------------------------ Date: 29 Oct 86 14:37:31 GMT From: ihnp4!houxm!mtuxo!mtune!mtunf!mtx5c!mtx5d!mtx5a!mtx5e!mtx5w!drv@ ucbvax.Berkeley.EDU Subject: Re: OPINIONS REQUESTED ON GURU > I'D APPRECIATE ANY COMMENTS THE GROUP HAS ON THE AI BASED PACKAGE . > I had an evaluation copy of Guru here about a month ago. I found it an interesting package with a lot of nice features. I decided not to use it for a lot of reasons specific to my application but I'll try not to let them get in the way of my evaluation. First, a short description of what Guru has. In addition to a language and a set of features for creating rule-based systems, Guru contains a text editor, a spreadsheet, a communications package, a graphics package, a relational data base package, a Unix shell-like procedural language, a menu and user prompt facility and probably a few other things I've forgotten. The rule-based system, editor and spreadsheet are the parts I looked into most so my comments will be limited to those. The editor and spreadsheet are not what you would call state-of-the-art. There are standalone packages available for most PCs that are as nice or nicer than Guru's in my opinion. While the menu interface to Guru and the graphics package make nice use of the PC graphics, neither the editor nor the spreadsheet use any graphics. It appears that the Guru folks purchased these packages from outside and integrated them in to their total system. That opinion is based on nothing other than the rather different appearance these modules have from each other. The novel and nice feature that Guru has that prompted my to look into it in the first place is the ability to reference different portions of Guru from others. For example, within a spreadsheet you can reference a rule-based system (which can access the data in the spreadsheet) and fill in cells with results from a rule- based execution (called a consultation in Guru). Similarly, within the editor you can access the data base for results to be added to the text, access the data base from within a rule based system, etc. I spent a fair amount of time with the spreadsheet accessing rules in a rule-based system. While I had a few difficulties due to the way the rules address spreadsheet cells, I found the procedure to work fairly well. One thing that turned me off from Guru, in addition to the mismatch with my intended application, was the price tag. $3000 seemed a bit steep for me. But if you need most or many of the different features rather than just a couple it might be a better investment instead of buying separate components. And if you need to have the integration between components such as spreadsheet and rule-based system, I know of no other tool that does that. Then the price might be well worth it. Good luck and I hope this helps. Dennis R. Vogel AT&T Information Systems Middletown, NJ (201) 957-4951 ------------------------------ Date: 31 Oct 86 15:43:16 GMT From: ihnp4!drutx!mtuxo!mtune!mtunf!mtx5c!mtx5d!mtx5a!mtx5e!mtx5w!drv@ ucbvax.Berkeley.EDU Subject: More on Guru I recently posted my experience with the Guru package. In it I mentioned that the $3000 price tag scared me off (in addition to other things). Well, it's worse than that. Today we got a letter from Guru saying that the introductory period for Guru has drawn to a close along with the $2995 introductory price. Guru is now priced at $6500 for a single user development system. I should mention that Guru does offer run-time licenses for less than this but the latest letter doesn't say what they cost. Dennis R. Vogel AT&T Information Systems Middletown, NJ (201) 957-4951 ------------------------------ Date: 30 Oct 86 23:13:47 GMT From: fluke!ssc-vax!bcsaic!michaelm@beaver.cs.washington.edu (Michael Maxwell) Subject: Re: Seeking PD parser for simple English sentences. In article <30@orion.UUCP> heins@orion.UUCP (Michael Heins) writes: >I am looking for public domain software which I can use to help me parse >simple English sentences into some kind of standardized representation. >I guess what I am looking for would be a kind of sentence diagrammer >which would not have to have any deep knowledge of the meanings of the >nouns, verbs, adjectives, etc. > >...C routines would be ideal. Also, references to published >algorithms would be useful. Since this seems to be a fairly common request, I am taking the liberty of posting to the net... Many Prologs (but not Turbo) have a built-in parser called `Definite Clause Grammar' (DCG). It is a way of writing phrase structure rules, which Prolog then translates into standard Prolog rules. Most standard texts on Prolog discuss it, e.g. %A W.F. Clocksin %A C.S. Mellish %D 1984 %T Programming in Prolog %I Springer-Verlag %C Berlin A somewhat more sophisticated rule system was developed by Fernando Pereira in his Ph.D. dissertation, published with some revision as: %A Fernando Pereira %D 1979 %T Extraposition Grammars %R Working Paper No. 59 %I Department of Aritficial Intelligence, University of Edinburgh %C Edinburgh (You'd have to type the program in yourself; he includes a very simple grammar of English.) -- Mike Maxwell Boeing Advanced Technology Center ...uw-beaver!uw-june!bcsaic!michaelm ------------------------------ Date: Mon, 27 Oct 86 16:47:16 EST From: "William J. Rapaport" Subject: KR for Music [Forwarded from the NL-KR Digest.] For information on KR and music, see: Ebcioglu, Kemal, "An Expert System for Chorale Harmonization," Proc. AAAI-86, Vol. 2, pp. 784-788. Ebcioglu, Kemal, "An Expert System for Harmonization of Chorales in the Style of J. S. Bach," Tech. Report, Dept. of Computer Science, SUNY Buffalo (1986). ------------------------------ Date: Fri, 31 Oct 86 15:54:18 est From: lb0q@andrew.cmu.edu (Leslie Burkholder) Subject: monotonicity The monotonicity property of validity: If an argument is deductively valid then it cannot be made invalid by adding new premises. Equivalently: If X, Y are finite sets of sentences and S a sentence, then if X entails S, then X union Y entails S. The monotonicity property of consistency: If a set of sentences is inconsistent then it cannot be made consistent by adding to it a new sentence. Equivalently: If X is a finite set of sentences, S some sentence, and X is inconsistent, then so is X union {S}. Leslie Burkholder ------------------------------ Date: 2 Nov 86 04:04:36 GMT From: rutgers!clyde!watmath!watnot!watdcsu!brewster@seismo.css.gov (dave brewer, SD Eng, PAMI ) Subject: Weizenbaum keynote address at U of Waterloo (long) The Hagey Lectures at the University of Waterloo provide an opportunity for a distinguished researcher to address the community at large every year. This year, Dr. Weizenbaum of MIT was the chosen speaker, and he has just delivered two key note addresses entitled; "Prospects for AI" and "The Arms Race, Without Us". The important points of the first talk can be summarized as : 1) AI has good prospects from an investment prospective since a strong commitment to marketing something called AI has been made. 2) the early researchers did not understand how difficult the problems they addressed were and so the early claims of the possibilities were greatly exaggerated. The trend still continues but on a reduced scale. 3) AI has been a handle for some portion of the US military to hang SDI on, since whenever a "difficult" problem arises it is always possible to say , " Well, we don't understand that now, but we can use AI techniques to solve that problem later." 4) the actual achievements of AI are small. 5) the ability of expert systems to continuously monitor stock values and react has led to increased volatility and crisis situations in the stock markets of the world recently. What happens if machine induced technical trading drops the stock market by 20 % in one day , 50 % in one day ? The important points of the second talk can be summarized as : 1) not all problems can be reduced to computation, for example how could you conceive of coding the human emotion loneliness. 2) AI will never duplicate or replace human intelligence since every organism is a function of its history. 3) research can be divided into performance mode or theory mode research. An increasing percentage of research is now conducted in performance mode, despite possible desires to do theory mode research, since funds (mainly military), are available for performance mode research. 4) research on "mass murder machines" is possible because the researchers (he addressed computer scientists directly although extension to any technical or scientific discipline was implied), are able to psychologically distance themselves from the end use of their work. 5) technical education that neglects language, culture, and history, may need to be rethought. 6) courage is infectious, and while it may not seem to be a possibility to some, the arms race could be stopped cold if an entire group of professions, (ie computer scientists), refused to participate. 7) the search for funds has led to an increased rate of performance mode research, and has even induced many institutions to prostitute themselves to the highest bidder. Specific situations within MIT were used for examples. Weizenbaum had the graciousness to ignore related (albeit proportionally smaller), circumstances at this university. 8) every researcher should assess the possible end use of their own research, and if they are not morally comfortable with this end use, they should stop their research. Weizenbaum did not believe that this would be the end of all research, but if that was the case then he would except this result. He specifically referred to research in machine vision, which he felt would be used directly and immediately by the military for improving their killing machines. While not saying so, he implied that this line of AI should be stopped dead in its tracks. Poster's comments : 1) Weizenbaum seemed to be technically out of date in some areas, and admitted as much at one point. Some of his opinions regarding state of the art were suspect. 2) His background, technical and otherwise, seems to predispose him to dismissing some technical issues a priori. i.e. a machine can never duplicate a human, why ?, because !. 3) His most telling point, and one often ignored, is that researchers have to be responsible for their work, and should consider its possible end uses. 4) He did not appear to have thought through all the consequences of a sudden end to research, and indeed many of his solutions appear overly simplistic, in light of the complicated world we live in. 5) You have never seen an audience squirm, as they did for the second lecture. A once premier researcher, addresses his contemporaries, and tells them they are ethically and morally bankrupt, and every member of the audience has at least some small buried doubt that maybe he is right. 6) Weizenbaum intended the talks to be "controversial and provocative" and has achieved his goal within the U of W community. While not agreeing with many of his points, I believe that there are issues raised which are relevant to the entire world-wide scientific community, and have posted for this reason. The main question that I see arising from the talks is : is it time to consider banning, halting, slowing, or otherwise rethinking certain AI or technical adventures, such as machine vision, as was done in the area of recombinant DNA. Disclaimer : The opinions above are mine and may not accurately reflect those of U of Waterloo, Dr.Weizenbaum, or anyone else for that matter. I make no claims as to the accuracy of the above summarization and advise that transcripts of the talks are available from some place within U of W, but expect to pay for them because thats the recent trend. UUCP : {decvax|ihnp4}!watmath!watdcsu!brewster Else : Dave Brewer, (519) 886-6657 ------------------------------ Date: 3 Nov 86 02:48:23 GMT From: tektronix!reed!trost@ucbvax.Berkeley.EDU (Bill Trost) Subject: Re: Weizenbaum keynote address at U of Waterloo (long) In article <2689@watdcsu.UUCP> brewster@watdcsu.UUCP (dave brewer, SD Eng, PAMI ) writes: > >The main question that I see arising from the talks is : is it time >to consider banning, halting, slowing, or otherwise rethinking >certain AI or technical adventures, such as machine vision, as was >done in the area of recombinant DNA. Somehow, I don't think that banning machine vision makes much sense. It seems that it would be similar to banning automatic transmissions because you can use them to make tanks. The device itself is not the hazard (as it is in genetic research) -- it is the application. -- Bill Trost, tektronix!reed!trost "ACK!" (quoted, without permission, from Bloom County) ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Fri Nov 7 19:19:39 1986 Date: Fri, 7 Nov 86 19:19:29 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe Subject: AIList Digest V4 #245 Status: R AIList Digest Wednesday, 5 Nov 1986 Volume 4 : Issue 245 Today's Topics: Queries - Electronics Troubleshooting & Consciousness & English-Hangul Machine Translation & Belle, AI Tools - Object-Oriented Programming, Ethics - Moral Responsibility, Logic - Nonmonotonic Reasoning, Linguistics - Nonsense Tests and the Ignorance Principle, Philosophy - Artificial Humans & Mathematics and Humanity ---------------------------------------------------------------------- Date: 3 Nov 86 07:23 PST From: kendall.pasa@Xerox.COM Subject: Electronics Troubleshooting For AI Digest 11/3/86: I am looking for some public domain software that would pertain to electronics troubleshooting. I am interested in any special graphics editors for electronics schematics and any shells for troubleshooting the circuits. Thank you. ------------------------------ Date: Tue, 4 Nov 86 00:41 EDT From: JUDD%cs.umass.edu@CSNET-RELAY.ARPA Subject: a query for AIList I would like references to books or articles that argue the following position: "The enigma of consciousness is profound but the basis for it is very mundane. Consciousness reduces to sensation; sensation reduces to measurement of nervous events; nervous events are physical events. Since physical events can obviously occur in `inanimate matter', so can consciousness." References to very recent articles are my main objective here, but old works (ie pre-computer age) would be helpful also. sj ------------------------------ Date: Mon, 3 Nov 86 20:53:05 EST From: "Maj. Ken Rose" (USATRADOC | mort) Subject: English-Hangul Machine Translation I would like to connect with anyone who has had some experience with machine translation between English and Hangul. Please contact me at krose@brl.arpa or phone (804) 727-2347. May-oo kam s'ham nee da. ------------------------------ Date: 4 Nov 86 15:11 EST From: Vu.wbst@Xerox.COM Reply-to: Vu.wbst@Xerox.COM Subject: Belle Does anyone have any information about Belle, the strongest chess-playing program today? It has ELO rating of over 2000. It uses tree-pruning, and cutoff heuristics. Is the code available for the public in Interlisp-D? Any pointer would be very helpful. By the way, what is the complete path for net.games.chess that was mentioned in V4# 243? Thank you. Dinh Vu ------------------------------ Date: Mon, 3 Nov 86 12:25:40 -0100 From: Hakon Styri Subject: Re: Is there OOP in AI? In response to the item in AIList issue No. 231. Yes, if you have a look in the October issue of SIGPLAN NOTICES, i.e. the special issue on the Object-Oriented Programming Workshop at IBM Yorktown Heights, June 1986. At least two papers will be of interest... Going back a few years, you may also find some ICOT papers about OOP in Prolog. Try New Generation Computing, the ICOT/Springer-Verlag Journal. There are a few papers in Vol. 1, No. 1 (1983). In the Proceedings of the International Conference on FGCS 1984 there is another paper: "Unique Features of ESP", which is a Logic Programming Language with features for OOP. H. Styri -- Yu No Hoo :-) ------------------------------ Date: Mon 3 Nov 86 10:59:37-PST From: cas Subject: moral responsibility The idea of banning vision research ( or any other, for that matter ) is even sillier and more dangerous than Bill Trost points out. The analogy is not to ban automatic transmissions, but THINKING about automatic transmissions. And banning thinking about anything is about as dangerous as any course of action can be, no matter how highminded or sincerely morally concerned those who call for it. To be fair to Weizenbaum, he does have a certain weird consistency. He tells me, for example, that in his view helicopters are intrinsically evil ( as the Vietnam war has shown ). One can see how the logic works: if an artifact is ( or can be ) used to do more bad than good, then it is evil, and research on evil things is immoral. While this is probably not the place to start a debate in theoretical ethics, I do think that this view, while superficially attractive, simply doesnt stand up to a little thought, and can be used to label as wicked anything which one dislikes for any reason at all. Weizenbaum has made a successful career by systematically attacking AI research on the grounds that it is somehow immoral, and finding a large and willing audience. He doesnt make me squirm. Pat Hayes ------------------------------ Date: Mon, 3 Nov 86 11:09 ??? From: GODDEN%gmr.com@CSNET-RELAY.ARPA Subject: banning machine vision As regards the banning of research on machine vision (or any technical field) because of the possible end uses of such technology in "mass murder machines", I should like to make one relevant distinction. The immediate purpose of the research as indicated in the grant proposal or as implied by the source of funding is of the utmost importance. If I were to do research on vision for the auto industry and published my results which some military type then schlepped for use in a satellite to zap civilians, I would not feel ANY responsibility if the satellite ever got used. On the other hand, if I worked for a defense contractor doing the same research, I certainly would bear some responsibility for its end use. -Kurt Godden godden@gmr.com ------------------------------ Date: Sat, 1 Nov 86 16:58:21 pst From: John B. Nagle Reply-to: jbn@glacier.UUCP (John B. Nagle) Subject: Re: Non-monotonic Reasoning Proper mathematical logic is very "brittle", in that two axioms that contradict each other make it possible to prove TRUE=FALSE, from which one can then prove anything. Thus, AI systems that use traditional logic should contain mechanisms to prevent the introduction of new axioms that contradict ones already present; this is referred to as "truth maintenance". Systems that lack such mechanisms are prone to serious errors, even when reasoning about things which are not even vaguely related to the contradictory axioms; one contradiction in the axioms generally destroys the system's ability to get useful results. Non-monotonic reasoning is an attempt to make reasoning systems less brittle, by containing the damage that can be caused by contradiction in the axioms. The rules of inference of non-monotonic reasoning systems are weaker than those of traditional logic. There is not full agreement on what the rules of inference should be in such systems. There are those who regard non-monotonic reasoning as hacking at the mathematical logic level. Non-monotonic reasoning lies in a grey area between the worlds of logic and heuristics. John Nagle ------------------------------ Date: Tue, 4 Nov 86 09:30:29 CST From: mklein@aisung.cs.uiuc.edu (Mark Klein) Subject: Nonmonotonic Reasoning My understanding of nonmonotonic reasoning (NMR) is different from what you described in a recent ailist posting. As I understand it, NMR differs from monotonic reasoning in that the size of the set of true theorems can DECREASE when you add new axioms - thus the set size does not increase monotonically as axioms are added. It is often implemented using truth maintenance systems (TMS) that allow something to be justified by something else NOT being believed. Monotonic reasoning, by contrast, could be implemented by a TMS that only allows something to be justified by something else being believed. Default reasoning is an instance of nonmonotonic reasoning. Nonmonotonic reasoning is thus not synonymous with truth maintenance. Mark Klein ------------------------------ Date: Thu, 30 Oct 86 17:26:02 est From: Graeme Hirst Subject: Re: Nonsense tests and the Ignorance Principle >A couple of years ago, on either this list or Human-Nets, there appeared >a short multiple-choice test which was written so that one could deduce >"best" answers based on just the form, not the content, of the questions ... The Ignorance Principle ("Forget the content, just use the form", or "Why clutter up the system with a lot of knowledge?") has a long and distinguished history in AI. In 1975, Gord McCalla and Michael Kuttner published an article on a program that could successfully pass the written section of the British Columbia driver's license exam. For example: QUESTION: Must every trailer which, owing to size or construction tends to prevent a driving signal given by the driver of the towing vehicle from being seen by the driver of the overtaking vehicle be equipped with an approved mechanical or electrical signalling device controlled by the driver of the towing vehicle? ANSWER: Yes. In fact, the program was able to answer more than half the questions just by looking for the words "must", "should", "may", "necessary", "permissible", "distance", and "required". The system was not without its flaws. For example: QUESTION: To what must the coupling device between a motor-vehicle and trailer be affixed? ANSWER: Yes. This is wrong; the correct answer is "frame". (This was an early instance of the frame problem.) The authors subsequently attempted a similar program for defending PhD theses, but the results were never published. REFERENCE McCalla, G. and Kuttner, M. "An extensible feature-based procedural question answering system to handle the written section of the British Columbia driver's examination". _CSCSI/SCEIO Newsletter_ [now published as _Canadian Artificial Intelligence_], 1(2), February 1975, 59-67. \\\\ Graeme Hirst University of Toronto Computer Science Department //// utcsri!utai!gh / gh@ai.toronto.edu / 416-978-8747 ------------------------------ Date: 27 Oct 86 09:38 PST From: Ghenis.pasa@Xerox.COM Subject: Why we waste time training machines >why are we wasting time training machines when we could be training humans >instead. The only reasons that I can see are that intelligent systems can be >made small enough and light enough to sit on bombs. Are there any other reasons? Why do we record music instead of teaching everyone how to sing? To preserve what we consider top performance and make it easily available for others to enjoy, even if the performer himself cannot be present and others are not inclined to or capable of duplicating his work, but simply wish to benefit from it. In the case of applied AI there is the added advantage that the "recordings" are not static but extendable, so the above question may be viewed as a variation of "to stand on the shoulders of giants" vs. "to reinvent the wheel". This is just ONE of the reasons we "waste our time" the way we do. -- Pablo Ghenis, speaking for myself (TO myself most of the time) ------------------------------ Date: Mon, 27 Oct 86 17:29 CDT From: stair surfing - an exercise in oblivion <"NGSTL1::EVANS%ti-eg.csnet"@CSNET-RELAY.ARPA> Subject: reasons for making "artificial humans" >In the last AI digest (V4 #226), Daniel Simon writes: > >>One question you haven't addressed is the relationship between intelligence and >>"human performance". Are the two synonymous? If so, why bother to make >>artificial humans when making natural ones is so much easier (not to mention >>more fun)? > >This is a question that has been bothering me for a while. When it is so much >cheaper (and possible now, while true machine intelligence may be just a dream) >why are we wasting time training machines when we could be training humans in- >stead. The only reasons that I can see are that intelligent systems can be made >small enough and light enough to sit on bombs. Are there any other reasons? > >Daniel Paul > >danny%ngstl1%ti-eg@csnet-relay First of all, I'd just like to comment that making natural humans may be easier (and more fun) for men, but it's not necessarily so for women. It also seems that once we get the procedure for "making artificial humans" down pat, it would take less time and effort than making "natural" ones, a process which currently requires at least twenty years (sometimes more or less). Now to my real point - I can't see how training machines could be considered a waste of time. There are thousands of useful but meaningless (and generally menial) jobs which machines could do, freeing humans for more interesting pursuits (making more humans, perhaps). Of more immediate concern, there are many jobs of high risk - mining, construction work, deep-sea exploration and so forth - in which machines, particularly intelligent machines, could assist. Putting intelligent systems on bombs is a minor use, of immediate concern only for its funding potentials. Debating the ethics of such use is a legitimate topic, I suppose, but condemning all AI research on that basis is not. Eleanor Evans evans%ngstl1%ti-eg@csnet-relay ------------------------------ Date: 28 Oct 86 17:18:00 EST From: walter roberson Subject: Mathematics, Humanity Gilbert Cockton recently wrote: >This is contentious and smacks of modelling all learning procedures >in terms of a single subject, i.e. mathematics. I can't think of a >more horrible subject to model human understanding on, given the >inhumanity of most mathematics! The inhumanity of *most* mathematics? I would think that from the rest of your message, what you would really claim is the inhumanity of *all* mathematics -- for *all* of mathematics is entirely deviod of the questions of what is morally right or morally wrong, entirely missing all matters of human relationships. Mathematical theorems start by listing the assumptions, and then indicating how those assumptions imply a result. Many humans seem to devote their entire lifes to forcibly changing other people's assumptions (and not always for the better!); most people don't seem to care about this process. Mathematics, then, could be said to be the study of single points, where "real life" requires that humans be able to adapt to a line or perhaps something even higher order.| And yet that does not render mathematics "inhumane", for we humans must always react to the single point that is "now", and we *do* employ mathematics to guide us in that reaction. Thus, mathematics is not inhumane at all -- at worst, it is a subclass of "humanity". If you prefer to think if it in such terms, this might be expressed as " !! Humanity encompasses something Universal!" Perhaps, though, there should be a category of study devoted to modelling the transformation of knowledge as the very assumptions change. A difficult question, of course, is whether such a study should attempt to, in any way, model the "morality" of changing assumptions. I would venture that it should not, but that a formal method of measuring the effects of such changes would not be out of order. ----- Gilbert, as far as I can tell, you have not presented anything new in your article. Unless I misunderstand you completely, your entire arguement is based upon the premise that there is something special about life that negates the possibility of life being modelled by any formal system, no matter how complex. As I personally consider that it might be possible to do such a modelling note that I don't say that it *is* possible to do such a modelling|, I disregard the entire body of your arguements. The false premise implies all conclusions. ----- >Nearer to home, find me >one computer programmer who's understanding is based 100% on formal procedures. >Even the most formal programmers will be lucky to be in program-proving mode >more than 60% of the time. So I take it that they don't `understand' what >they're doing the other 40% of the time? I'm not quite sure what you mean to imply by "program-proving mode". The common use of the word "prove" would imply "a process of logically demonstrating that an already-written program is correct". The older use of "prove" would imply "a process of attempting to demonstrate that an already- written program is incorrect." In either case, the most formal of programmers spend relatively little time in "program-proving mode", as those programmers employ formal systems to write programs which are correct in the first place. It is only those that either do not understand programming, or do not understand all the implications of the assumptions they have programmed, that require 60% of their time to "prove" their programs. 60% of their time proving to others the validity of the approach, perhaps... walter roberson walter ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Fri Nov 7 19:18:25 1986 Date: Fri, 7 Nov 86 19:18:16 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #246 Status: R AIList Digest Wednesday, 5 Nov 1986 Volume 4 : Issue 246 Today's Topics: Philosophy - The Analog/Digital Distinction ---------------------------------------------------------------------- Date: 29 Oct 86 17:34:08 GMT From: rutgers!princeton!mind!harnad@lll-crg.arpa (Stevan Harnad) Subject: Re: The Analog/Digital Distinction: Sol Concerning the A/D distinction, goldfain@uiucuxe.CSO.UIUC.EDU replies: > Analog devices/processes are best viewed as having a continuous possible > range of values. (An interval of the real line, for example.) > Digital devices/processes are best viewed as having an underlying > granularity of discrete possible values. > (Representable by a subset of the integers.) > This is a pretty good definition, whether you like it or not. > I am curious as to what kind of discussion you are hoping to get, > when you rule out the correct distinction at the outset ... Nothing is ruled out. If you follow the ongoing discussion, you'll see what I meant by continuity and discreteness being "nonstarters." There seem to be some basic problems with what these mean in the real physical world. Where do you find formal continuity in physical devices? And if it's only "approximate" continuity, then how is the "exact/approximate" distinction that some are proposing for A/D going to work? I'm not ruling out that these problems may be resolvable, and that continuous/discrete will emerge as a coherent criterion after all. I'm just suggesting that there are prima facie reasons for thinking that the distinction has not yet been formulated coherently by anyone. And I'm predicting that the discussion will be surprising, even to those who thought they had a good, crisp, rigorous idea of what the A/D distinction was. Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ Date: 29 Oct 86 16:37:55 GMT From: rutgers!husc6!Diamond!aweinste@lll-crg.arpa (Anders Weinstein) Subject: Re: The Analog/Digital Distinction > From Stevan Harnad: > >> Analog signal -- one that is continuous both in time and amplitude. >> ... >> Digital signal -- one that is discrete both in time and amplitude... >> This is obtained by quantizing a sampled signal. > > Question: What if the >original "object" is discrete in the first place, both in space and >time? Does that make a digital transformation of it "analog"? I Engineers are of course free to use the words "analog" and "digital" in their own way. However, I think that from a philosophical standpoint, no signal should be regarded as INTRINSICALLY analog or digital; the distinction depends crucially on how the signal in question functions in a representational system. If a continuous signal is used to encode digital data, the system ought to be regarded as digital. I believe this is the case in MOST real digital systems, where quantum mechanics is not relevant and the physical signals in question are best understood as continuous ones. The actual signals are only approximated by discontinous mathematical functions (e.g. a square wave). > The image of an object >(or of the analog image of an object) under a digital transformation >is "approximate" rather than "exact." What is the difference between >"approximate" and "exact"? Here I would like to interject a tentative >candidate criterion of my own: I think it may have something to do with >invertibility. A transformation from object to image is analog if (or >>to the degree that) it is invertible. In a digital approximation, some >information or structure is irretrievably lost (the transformation >is not 1:1). > ... It's a mistake to assume that transformation from "continuous" to "discrete" representations necessarily involves a loss of information. Lots of continuous functions can be represented EXACTLY in digital form, by, for example, encoded polynomials, differential equations, etc. Anders Weinstein ------------------------------ Date: 29 Oct 86 20:28:06 GMT From: rutgers!princeton!mind!harnad@titan.arc.nasa.gov (Stevan Harnad) Subject: Re: The Analog/Digital Distinction [Will someone with access post this on sci.electronics too, please?] Anders Weinstein has offered some interesting excerpts from the philosopher Nelson Goodman's work on the A/D distinction. I suspect that some people will find Goodman's considerations a little "dense," not to say hirsute, particularly those hailing from, say, sci.electronics; I do too. One of the subthemes here is whether or not engineers, cognitive psychologists and philosophers are talking about the same thing when they talk about A/D. [Other relevant sources on A/D are Zenon Pylyshyn's book "Computation and Cognition," John Haugeland's "Artificial Intelligence" and David Lewis's 1971 article in Nous 5: 321-327, entitled "Analog and Digital."] First, some responses to Weinstein/Goodman on A/D; then some responses to Weinstein-on-Harnad-on-Jacobs: > systems like musical notation which are used to DEFINE a work of > art by dividing the instances from the non-instances I'd be reluctant to try to base a rigorous A/D distinction on the ability to make THAT anterior distinction! > "finitely differentiated," or "articulate." For every two characters > K and K' and every mark m that does not belong to both, [the] > determination that m does not belong to K or that m does not belong > to K' is theoretically possible. ... I'm skeptical that the A/D problem is perspicuously viewed as one of notation, with, roughly, (1) the "digital notation" being all-or-none and discrete and the "analog notation" failing to be, and with (2) corresponding capacity or incapacity to discriminate among the objects they stand for. > A scheme is syntactically dense if it provides for infinitely many > characters so ordered that between each two there is a third. I'm no mathematician, but it seems to me that this is not strong enough for the continuity of the real number line. The rational numbers are "syntactically dense" according to this definition. But maybe you don't want real continuity...? > semantic finite differentiation... for every two characters > I and K' such that their compliance classes are not identical and [for] > every object h that does not comply with both, [the] determination > that h does not comply with K or that h does not comply with K' must > be theoretically possible. I hesitantly infer that the "semantics" concerns the relation between the notational "image" (be it analog or digital) and the object it stands for. (Could a distinction that so many people feel they have a good intuitive handle on really require so much technical machinery to set up? And are the different candidate technical formulations really equivalent, and capturing the same intuitions and practices?) > A symbol _scheme_ is analog if syntactically dense; a _system_ is > analog if syntactically and semantically dense. ... A digital scheme, > in contrast, is discontinuous throughout; and in a digital system the > characters of such a scheme are one-one correlated with > compliance-classes of a similarly discontinous set. But discontinuity, > though implied by, does not imply differentiation...To be digital, a > system must be not merely discontinuous but _differentiated_ > throughout, syntactically and semantically... Does anyone who understands this know whether it conforms to, say, analog/sampled/quantized/digital distinctions offered by Steven Jacobs in a prior iteration? Or the countability criterion suggested by Mitch Sundt? > If only thoroughly dense systems are analog, and only thoroughly > differentiated ones are digital, many systems are of neither type. How many? And which ones? And where does that leave us with our distinction? Weinstein's summary: >>To summarize: when a dense language is used to represent a dense domain, the >>system is analog; when a discrete (Goodman's "discontinuous") and articulate >>language maps a discrete and articulate domain, the system is digital. What about when a discrete language is used to represent a dense domain (the more common case, I believe)? Or the problem case of a dense representation of a discrete domain? And what if there are no dense domains (in physical nature)? What if even the dense/dense criterion can never be met? Is this all just APPROXIMATELY true? Then how does that square with, say, Steve Jacobs again, on approximation? -------- What follows is a response to Weinstein-on-Harnad-on-Jacobs: > Engineers are of course free to use the words "analog" and "digital" > in their own way. However, I think that from a philosophical > standpoint, no signal should be regarded as INTRINSICALLY analog > or digital; the distinction depends crucially on how the signal in > question functions in a representational system. If a continuous signal > is used to encode digital data, the system ought to be regarded as > digital. Agreed that an isolated signal's A or D status cannot be assigned, and that it depends on its relation with other signals in the "representational system" (whatever that is) and their relations to their sources. It also depends, I should think, on what PROPERTIES of the signal are carrying the information, and what properties of the source are being preserved in the signal. If the signal is continuous, but its continuity is not doing any work (has no signal value, so to speak), then it is irrelevant. In practice this should not be a problem, since continuity depends on a signal's relation to the rest of the signal set. (If the only amplitudes transmitted are either very high or very low, with nothing in between, then the continuity in between is beside the point.) Similarly with the source: It may be continuous, but the continuity may not be preserved, even by a continuous signal (the continuities may not correlate in the right way). On the other hand, I would want to leave open the question of whether or not discrete sources can have analogs. > I believe this is the case in MOST real digital systems, where > quantum mechanics is not relevant and the physical signals in > question are best understood as continuous ones. The actual signals > are only approximated by discontinous mathematical functions (e.g. > a square wave). There seems to be a lot of ambiguity in the A/D discussion as to just what is an approximation of what. On one view, a digital representation is a discrete approximation to a continuous object (source) or to a (continuous) analog representation of a (continuous) object (source). But if all objects/sources are really discontinuous, then it's really the continuous analog representation that's approximate! Perhaps it's all a matter of scale, but then that would make the A/D distinction very relative and scale-dependent. > It's a mistake to assume that transformation from "continuous" to > "discrete" representations necessarily involves a loss of information. > Lots of continuous functions can be represented EXACTLY in digital > form, by, for example, encoded polynomials, differential equations, etc. The relation between physical implementations and (formal!) mathematical idealizations also looms large in this discussion. I do not, for example, understand how you can represent continuous functions digitally AND exactly. I always thought it had to be done by finite difference equations, hence only approximately. Nor can a digital computer do real integration, only finite summation. Now the physical question is, can even an ANALOG computer be said to be doing true integration if physical processes are really discrete, or is it only doing an approximation too? The only way I can imagine transforming continuous sources into discrete signals is if the original continuity was never true mathematical continuity in the first place. (After all, the mathematical notion of an unextended "point," which underlies the concept of formal continuity, is surely an idealization, as are many of the infinitesmal and limiting notions of analysis.) The A/D distinction seems to be dissolving in the face of all of these awkward details... Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Fri Nov 7 19:18:41 1986 Date: Fri, 7 Nov 86 19:18:31 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #247 Status: R AIList Digest Wednesday, 5 Nov 1986 Volume 4 : Issue 247 Today's Topics: Philosophy - Defining the Analog/Digital Distinction ---------------------------------------------------------------------- Date: 30 Oct 86 02:16:42 GMT From: rutgers!princeton!mind!harnad@SPAM.ISTC.SRI.COM (Stevan Harnad) Subject: Re: Defining the Analog/Digital Distinction [Could someone post this to sci.electronics, to which I have no access, please?] ------ (1) ken@rochester.arpa writes: > I think the distinction is simply this: digital deals with a finite set > of discrete {voltage, current, whatever} levels, while analog deals > with a *potentially* infinite set of levels. Now I know you are going > to say that analog is discrete at the electron noise level but the > circuits are built on the assumption that the spectrum is continuous. > This leads to different mathematical analyses. It sounds as if a problem of fact is being remedied by an assumption here. Nor do potential infinities appear to remedy the problem; there are perfectly discrete potential infinities. The A/D distinction is again looking approximate, relative and scale-dependent, hence, in a sense, arbitrary. > Sort of like infinite memory Turing machines, we don't have them but > we program computers as if they had infinite memory and in practice > as long as we don't run out, it's ok. So as long as we don't notice > the noise in analog, it serves. An approximation to an infinite rote memory represents no problem of principle in computing theory and practice. But an approximation to an exact distinction between the "exact" and the "approximate" doesn't seem satisfactory. If there is an exact distinction underlying actual engineering practice, at least, it would be useful to know what it was, in place of intuitions that appear to break down as soon as they are made precise. -------- (2) cuuxb!mwm (Marc Mengel) writes: > Digital is essentially a subset of analog, where the range of > properties used to represent information is grouped into a > finite set of values... > Analog, on the other hand, refers to using a property to directly > represent an infinite range of values with a different infinite > range of values. This sounds right, as far as it goes. D may indeed be a subset of A. To use the object--transformation--image vocabulary again: When an object is transformed into an image with only finite values, then the transform is digital. (What about combinations of image values?) When an infinite-valued object is transformed into an infitinite-valued (and presumably covariant) image, then the transform is analog. I assume the infinities in question have the right cardinality (i.e., uncountable). Questions: (i) Do discrete objects, with only finite or countably infinite properties, not qualify to have analogs? (ii) What does "directly represent" mean? Is there something indirect about finiteness? (iii) What if there are really no such infinities, physically, on either the object end or the image end? May I interject at this point the conjecture that what seems to be left out of all these A/D considerations so far (not just this module) is that discretization is usually not the sole means or end of digital representation. What about symbolic representation? What turns a discretized, approximate image of an object into a symbolic representation, manipulable by formal rules and semantically interpretable as being a representation OF that object? (But perhaps this is getting a little ahead of ourselves.) > This is why slide-rules are considered analog, you are USING distance > rather than voltage, but you can INTERPRET a distance as precisely > as you want. An abacus, on the otherhand also USES distance, but > where a disk is MEANS either one thing or another, and it takes > lots of disks to REPRESENT a number. An abacus then, is digital. (No comment. Upper case added.) -------- (3) writes: > (An) analog is a (partial) DUPLICATE (or abstraction) > of some material thing or some process, which contains > (it is hoped) the significant characteristics and properties > of the original. And a digital representation can't be any of these things? "Duplicate" in what sense? An object's only "exact" double is itself. Once we move off in time and space and properties, more precise notions of "duplicate" are needed than the intuitive ones. Sharing the SAME physical properties (e.g., obeying the same differential equations [thanks to Si Kochen for that criterion])? Or perhaps just ANALOGS of them? But then that gets a bit circular. > A digital device or method operates on symbols, rather than > physical (or other) reality. Analog computers may operate on > (real) voltages and electron flow, while digital computers > operate on symbols and their logical interrelationships. On the face of it, digital computers "operate" on the same physical properties and principles that other physical mechanisms do. What is different is that some aspects of their operations are INTERPRETABLE in special ways, namely, as rule-governed operations of symbol tokens that STAND FOR something else. One of the burdens of this discussion is to determine precisely what role the A/D distinction plays in that phenomenon, and vice versa. What, to start with, is a symbol? > Digital operations are formal; that is they treat form rather > than content, and are therefore always deductive, while the > behavior of real things and their analogs is not. Unfortunately, however, these observations are themselves a bit too informal. What is it to treat form rather than content? One candidate that's in the air is that it is to manipulate symbols according to certain formal rules that indicate what to do with the symbol tokens on the basis of their physical shapes only, rather than what the tokens or their manipulations or combinations "stand for" or "mean." It's not clear that this definition is synonymous with symbol manipulation's always being "deductive." Perhaps it's interpretable as performing deductions, but as for BEING deductions, that's another question. And how can digital operations stand in contrast to the behavior of "real things"? Aren't computers real things? > It is one of my (unpopular) assertions that the central nervous > system of living organisms (including myself) is best understood > as an analog of "reality"; that most interesting behavior > such as induction and the detection of similarity (analogy and > metaphor) cannot be accomplished with only symbolic, and > therefore deductive, methods. Such a conjecture would have to be supported not only by a clear definition of all of the ambiguous theoretical concepts used (including "analog"), but by reasons and evidence. On the face of it, various symbol-manipulating devices in AI do do "induction" and "similarity detection." As to the role of analog representation in the brain: Perhaps we'd better come up with a viable literal formulation of the A/D distinction; otherwise we will be restricted to figurative assertions. (Talking too long about the analog tends to make one lapse into analogy.) -------- (4) lanl!a.LANL.ARPA!crs (Charlie Sorsby) writes: > It seems to me that the terms as they are *usually* used today > are rather bastardized... when the two terms originated they referred > to two ways of "computing" and *not* to kinds of circuits at all. > The analog simulator (or, more popularly, analog computer) "computed" > by analogy. And, old timers may recall, they weren't all electronic > or even electrical. But what does "compute by analogy" mean? > Digital computers (truly so) on the other hand computed with > *digits* (i.e. numbers). Of course there was (is) analogy involved > here too but that was a "higher-order term" in the view and was > conveniently ignored as higher order terms often are. What is a "higher-order term"? And what's the difference between a number and a symbol that's interpretable as a number? That sounds like a "higher-order" consideration too. > In the course of time, the term analog came to be used for those > electronic circuits *like* those used in analog simulators (i.e. > circuits that work with continuous quantities). And, of course, > digital came to refer to those circuits *like* those used in digital > computers (i.e. those which work with discrete or quantized quantities. You guessed my next question: What does "like" mean, and why does the underlying distinction correlate with continuous and discrete circuit properties? > Whether a quantity is continuous or discrete depends on such things > as the attribute considered, to say nothing of the person doing the > considering, hence the vagueness of definition and usage of the > terms. This vagueness seems to have worsened with the passage of time. I couldn't agree more. And an attempt to remedy that is one of the objects of this exercise. -------- (5) sundt@mitre.ARPA writes: > Coming from a heavily theoretical undergraduate physics background, > it seems obvious that the ONLY distinction between the analog and > digital representation is the enumerability of the relationships > under the given representation. > First of all, the form of digital representation must be split into > two categories, that of a finite representation, and that of a > countably infinite representation. Turing machines assume a countably > infinite representation, whereas any physically realizable digital > computer must inherently assume a finite digital representation. > Second, there must be some predicate O(a,b) defined over all the a > and b in the representation such that the predicate O(a,b) yields > only one of a finite set of symbols, S(i) (e.g. "True/False"). > If such a predicate does not exist, then the representation is > arguably ambiguous and the symbols are "meaningless". > Looking at all the (a,b) pairs that map the O(a,b) predicate into > the individual S(i): > ANALOG REPRESENTATION: the (a,b) pairs cannot be enumerated for ALL S(i) > COUNTABLY-INFINITE DIGITAL REPRESENTATION: the (a,b) pairs cannot be > enumerated for ALL S(i). > FINITE DIGITAL REPRESENTATION: all the (a,b) pairs for all the S(i) > CAN be enumerated. > This distinguishes the finite digital representation from the other two > representations. I believe this is the distinction you were asking > about. The distinction between the analog representation and the > countably-infinite digital representation is harder to identify. > I sense it would require the definition of a mapping M(a,b) onto the > representation itself, and the study of how this mapping relates to > the O(a,b) predicate. That is, is there some relationship between > O(?,?), M(?,?) and the (a,b) that is analgous to divisibility in > Z and R. How this would be formulated escapes me. You seem to have here a viable formal definition of something that can be called a "analog representation," based on the formal notion of continuity and nondenumerability. The question seems to remain, however, whether it is indeed THIS precise sense of analog that engineers, cognitive psychologists and philosophers are informally committed to, and, if so, whether it is indeed physically realizable. It would be an odd sort of representation if it were only an unimplementable abstraction. (Let me repeat that the finiteness of physical computers is NOT an analogous impediment for turing-machine theory, because the finite approximations continue to make sense, whereas both the finite and the denumerably infinite approximation to the A/D distinction seems to vitiate the distinction.) It's not clear, by the way, that it wasn't in fact the (missing) distinction between a countable and an uncountable "representation" that would have filled the bill. But I'll assume, as you do, that some suitable formal abstraction would capture it. THe question remains: Does that capture our A/D intuitions too? And does it sort out all actual (physical) A/D cases correctly? -------- The rest of Mitch Sundt's reply pertains also to the "Searle, Turing, Categories, Symbols" discussion that is going on in parallel with this one: > we can characterize when something is NOT intelligent, > but are unable to define when it is. I don't see at all why this is true, apart from the fact that confirming or supporting an affirmation is always more open-ended than confirming or supporting a denial. > [Analogously] Any attempt to ["define chaos"] would give it a fixed > structure, and therefore order... Thus, it is the quality that > is lost when a signal is digitized to either a finite or a > countably-infinite digital representation. Analog representations > would not suffer this loss of chaos. Maybe they wouldn't, if they existed as you defined them, and if chaos were worth preserving. But I'm beginning to sense a gradual departure from the precision of your earlier formal abstractions in the direction of metaphor here... > Carrying this thought back to "intelligence," intelligence is the > quality that is lost when the behavior is categorized among a set > of values. Thus, to detect intelligence, you must use analog > representations (and meta-representations). And I am forced to > conclude that the Turing test must always be inadequate in assessing > intelligence, and that you need to be an intelligent being to > *know* an intelligent being when you see one! I think we have now moved from equating "analog" with a precise (though not necessarily correct) formal notion to a rather free and subjective analogy. I hope it's clear that the word "conclude" here does not have quite the same deductive force it had in the earlier considerations. > Thinking about it further, I would argue, in view of what I just > said, that people are by construction only "faking" intelligence, > and that we have achieved a complexity whereby we can percieve *some* > of the chaos left by our crude categorizations (perhaps through > multiple categorizations of the same phenomena), and that this > perception itself gives us the appearance of intelligence. Our > perceptions reveal only the tip of the chaotic iceberg, however, > by definition. To have true intelligence would require the > perception of *ALL* the chaos. Thinking too much about the mind/body problem will do that to you sometimes. Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Fri Nov 7 19:20:08 1986 Date: Fri, 7 Nov 86 19:19:44 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe Subject: AIList Digest V4 #248 Status: R AIList Digest Wednesday, 5 Nov 1986 Volume 4 : Issue 248 Today's Topics: Philosophy - The Analog/Digital Distinction ---------------------------------------------------------------------- Date: 3 Nov 86 05:37:31 GMT From: allegra!princeton!mind!harnad@ucbvax.Berkeley.EDU (Stevan Harnad) Subject: Analog/Digital Distinction: 8 more replies Here are 8 more contributions to the A/D Distinction from (1) M. Derthick, (2) B. Garton, (3) W. Hamscher, (4) D. B. Plate, (5) R. Thau, (6) B. Kuszmaul, (7) C. Timar and (8) A. Weinstein. My comments follow the excerpts from each: ----- (1) (mad@g.cs.cmu.edu> (Mark Derthick) writes: > John Haugeland uses digital (and discrete) to mean "perfectly > definite," which is, I think, the best that can be done. Thus > representing an integer as the length of a stick in inches is digital, > but using the length of the stick in angstroms isn't. Obviously there > is a fuzzy boundary between the two. By the way, it is no problem that > sticks can be 6.5" long, as long as there can be unambiguous cases. Unfortunately, it is just this fuzzy boundary that is at issue here. ----- (2) winnie!brad (Brad Garton) writes: > A couple [of] items you mentioned in passing about the a/d issue struck > some resonant chords in my mind. You hit the nail on the head for me > when you said something about the a/d distinction possibly being a > problem of scaling (I think you were replying to the idea of quantum > effects at some level). When I consider the digitized versions of analog > signals we deal with over here , it seems that we > approximate more and more closely the analog signal with the > digital one as we increase the sampling rate. This process reminds > me of Mandelbrot's original "How Long is the Coastline of Britain" > article dealing with fractals. Perhaps "analog" could be thought > of as the outer limit of some fractal set, with various "digital" > representations being inner cutoffs. Don't know how useful this > is, but I do like the idea of "analog" and "digital" being along > some sort of continuum. > You also posed a question about when an approximate image of something > becomes a symbol of that thing (please forgive my awful paraphrasing). > As you seemed to be hinting, this is indeed a pretty sticky and > problematic issue. It always amazes me how quickly people are able > to identify a sound as being artificial (synthesized) when the signal > was intended to simulate a 'natural' instrument, rather than when > the computer (or sunthesizer) was being used to explore some new > timbral realm. Context sensitive? (and I haven't even mentioned yet > the problems of signals in a "musical" phrase!). As you may have been noticing from the variety of the responses, the A/D distinction seems to look rather different from the standpoints of hardware concerns, signal analysis, computational theory, AI, robotics, cognitive modeling, physics, mathematics, philosophy and, evidently, music synthesis. And that's without mentioning poets and hermeneuts. My question about fractals would be similar to my question about continuity: Are they to be a LITERAL physical model? Or are they just an abstraction, as I believe the proposals based on true continuity and uncountability have so far been? Or are what I'm tentatively rejecting as "abstractions" in fact standard examples of nomological generalizations, like the ideal gas laws, perfect elasticity, frictionless planes, etc.? [I'm inclined to think they're not, because I don't think there is a valid counterpart, in the idealization of continuity in these A/D formulations, to the physical notions of friction, etc.. The latter are what account for why it is that we never observe the idealized pattern in physics (but only an approximation to it) and yet we (correctly) continue to take the idealizations to be literally true of nature.] ----- (3) hamscher@HT.AI.MIT.EDU (Walter Hamscher) replied as follows in mod.ai: > I don't read all the messages on AiList, so I may have missed > something here: but isn't ``analog vs digital'' the same thing as > ``continuous vs discrete''? Continuous vs discrete, in turn, can be > defined in terms of infinite vs finite partitionability. It's a > property of the measuring system, not a property of the thing being > measured. If you sample some of the other responses you'll see that some people think that something can be formally defined along those lines, but whether it is indeed the A/D Distinction remains to be seen. ----- (4) The next contribution, posted in sci.electronics by plate@dicome.UUCP (Douglas B. Plate) is somewhat more lyrical: > The complete workings of the universe are analog in nature, > the growth of a leaf, decay of atomic structures, passing of > electrons between atoms, etc. Analog is natural reality, > even though facts about it's properties may remain unknown, > the truth of ANALOG exists in an objective form. > DIGITAL is an invention, like mathematics. It is a representation, > and I will not make any asumptions about what it would represent > except that whatever it represents, being a part of this Universe, > would have the same properties and nature that all other things > in the Universe share. The goal of DIGITAL then would be to > represent things 100% accurately. I will not say that ANALOG is > an infinitely continuous process, because I cannot prove that > there is not a smallest possible element involved in an ANALOG > process, however taking observed phenomena into account, I would > risk to say that the smallest element of ANALOG have not been > measured yet if they do ideed exist. > Digital is finite only in the number of elements it uses to represent > and the practical problem is that "bits" would have to extend > into infinity or to a magnitude equalling the smallest element > of what ANALOG is made of, for digital to reach it's full potential. > The thing is, Analog has the "natural" advantage. The universe is > made of it and what is only theory to DIGITAL is reality to > ANALOG. The intrinsic goal of DIGITAL is to become like > ANALOG. Why? Because DIGITAL "represents" and until it > becomes like ANALOG in it's finity/infinity, all of it's > representions can only be approximation. > DIGITAL will forever be striving to attain what ANALOG > was "born with". In theory, DIGITAL is just as continuously > infinite as ANALOG, because an infinite number of bits could > be used to represent an infinite number of things with 100% > accuracy. In practice, ANALOG already has this "infinity" > factor built into it and DIGITAL, like a dog chasing it's own > tail, will be trying to catch up on into infinity. This personification of "the analog" and "the digital" certainly captures many peoples' intuitions, but unfortunately it remains entirely at the intuitive level. Anthony Wilden wrote a book along these lines that turned the analog and the digital into an undergraduate cult for a few years, very much the way the left-brain/right-brain has been. What I'm wondering whether this exercise can do is replace the hermeneutics by a coherent, explicit, empirical construct with predictive and explanatory power. ----- (5) On sci.math,sci.physics,sci.electronics rst@godot.think.com.UUCP (Robert Thau) replied as follows: > In article <105@mind.UUCP> harnad@mind.UUCP (Stevan Harnad) writes: >>"Preserving information under transformations" also sounds like a good >>candidate... I would think that the invertibility of analog >>transformations might be a better instance of information preservation than >>the irretrievable losses of A/D. > I'm not convinced. Common ways of transmitting analog signals all > *do* lose at least some of the signal, irretrievably. Stray > capacitance in an ordinary wire can distort a rapidly changing signal. > Even fiber optic cables lose signal amplitude enough to require > repeaters. Losses of information in processing analog signals tend to > be worse, and for an analog transformation to be exactly invertible, it > *must* preserve all the information in its input. But then wouldn't it be fairest to say that, to the degree that a signal FAILS to preserve its source's properties it is NOT an analog of it? > ...The point is that the amount > of information in the speakers' input which they lose, irretrievably, > is a consequence of the design decisions of the people who made them. > Such design decisions are as explicit as the number of bits used in a > digital representation of the signal in the CD player farther up the > pike. Either digital or analog systems can be made as "true" as you > like, given enough time, materials, and money, but in neither case is > perfection an option. But then what becomes of the often-proposed property of "approximateness" as a distinguisher of an analog representation from a digital one, if they're BOTH approximate? Thau closes by requoting me: >>And this still seems to side-step the question of WHAT information is >>preserved, and in what way, by analog and digital representations, >>respectively. to which he replies: > Agreed. I can't tell whether this is intended to be ironic or Thau is really acknowledging the residual burden of specifying what it is in the way the information is represented that makes it analog or digital, given that the approximate/exact distinction seems to fail and the continuous/discrete one seems to be either just an abstraction or fails to square with the physics. ----- (6) On sci.electronics,sci.physics,sci.math bradley@godot.think.com.UUCP (Bradley Kuszmaul) proposes the following very relativistic account: > The distinction between digital and analog is in our minds. > "digital" and "analog" are just names of design methodologies that > engineers use to build large systems. "digital" is not a property of > a signal, or a machine, but rather a property of the design of the > machine. The design of the machine may not be a part of the machine. > If I gave you a music box (which played music naturally), and > you might not be able to tell whether it was digital or analog (even > if you could open it up and look at it and probe various things with > oscilliscopes or other tools). > Suppose I gave you a set of schematics for the box in which > everything was described in terms of voltages and currents, and which > included an explanation of how the box worked using continuous > mathematical functions. The schematics might explain how various > subcomponents interpreted their inputs as real numbers (even though > the inputs might be a far cry from real numbers e.g. due to the > quantization of everything by physicists). You would probably > conclude that the music box was an analog device. > Suppose, on the other hand, that I gave you a set of schematics for > the same box in which all the subcomponents were described in terms of > discrete formulas (e.g. truth tables), and included an explanation of > how the inputs from reality are interpreted by the hardware as > discrete values (even though the inputs might be a far cry from > discrete values e.g. due to ``noise'' from the uncertainty of > everything). You would probably conclude that the music box was a > digital device. > The idea is that a "digital" designer and "analog" designer might > very well come up with the same hardware to solve some problem, but > they would just understand the behaviour differently. > If designers could handle the complexity of thinking about > everything, they would not use any of these abstractions, but would > just build hardware that works. Real designers, on the other hand, > must control the complexity of the systems they design, and the > "digital" and "analog" design methodologies control the complexity of > the design while preserving enough of reality to allow the engineer to > make progress. If I understand correctly, Kuszmaul is suggesting that whether a representation is analog or digital may just be a matter of interpretation. (This calls to mind some Searlian issues about "intrinsic" vs. "derived" intentionality.) I have some sympathy for this view, because I myself have had reason to propose that the very same "module" that is regarded as "digital" in its autonomous, stand-alone form, might be regarded as analogue in a "dedicated" system, with all inputs and outputs causally connected to the world, and hence all "interpretations" fixed. Of course, that's just a matter of scale too, since ALL systems, whether with or without human intermediaries and interpreters, are causally connected to the world... But this does do violence to whatever is guiding some people's intuitions on this, for they would claim that THEIR notion of analogue is completely interpretation-independent. The part of me that leans toward the invertibility/information-preserving criterion sides with them. > If you buy my idea that digital and analog literally are in our > minds, rather than in the hardware, then the problem is not one of > deciding whether some particular system is digital (such questions > would be considered ill-posed). The real problem, as I view it, is to > distinguish between the digital and analog design methodologies. > We can try to understand the difference by looking at the cases > where we would use one versus the other. > We often use digital systems when the answer we want is a number. > (such as the decimal expansion of PI to 1000 digits) > We often use analog systems when the answer we want is something > physical (I don't really have good examples. Many of the things > which were traditionally analog are going digital for some of the > reasons described below. e.g. music, pictures (still and moving), > the control of an automobile engine or the laundry machine) > Digital components are nice because they have specifications which > are relatively straightforward to test. To test an analog > component seems harder. Because they are easier to test, > they can be considered more "uniform" than analog components (a > TTL "OR" gate from one mfr is about the same as a TTL "OR" gate > from another). (The same argument goes the other way too...) > Analog components are nice because sometimes they do just what you > wanted. For example, the connection from the gas peddle to the > throttle on the carburator of a car can be made by a mechanical > linkage which gives output which is a (approximately) continuous > function of the input position. To "fly by wire" (i.e. to use a > digital linkage) requires a lot more technology. > (When I say "we use a digital system", I really mean that "we design > such a system using a digital methodology", and correspondingly for > the analog case) > There are of course all sorts of places between "digital" and > "analog". A system may have digital subsystems and analog subsystems > and there may be analog subsystems inside the digital subsystems and > it goes on and on. This sort of thing makes the decision about > whether some particular design methodology is digital or analog hard. I'll leave it to the A/D absolutists to defend against this extreme relativism. I still feel agnostic. Except I do believe that the system that will ultimately pass the Total Turing Test will be deeply hybrid through-and-through -- and not just a concatenation of add-on analog and digital modules either. ----- (7) Cary Timar writes: > A great deal of the problem with the definitions I've seen is a > vagueness in describing what continuous and discrete sets are. > The distinction does not lie in the size of the set. It is possible to > form a discrete set of arbitrary cardinality - the set of all ordinals > in the initial segment of the cardinal. This set will start with > 0,1,2,3,... which most people agree is discrete. > I would say that a space can be considered to be "discrete" if it is not > regular, and "continuous" if it is normal. I hesitate to classify the > spaces which are regular but not normal. Luckily, we seldom deal with > models of computation using values taken from such a space. > Actually, I should have looked all of this up before I mailed it, but > I'm getting lazy. If you want to try to find mathematical definitions > of discrete and continuous spaces, I would suggest starting from texts > on Topology, especially Point-Set Topology. I wouldn't trust any one > text to give an universally agreed on definition either... Of course, if I believed it was just a matter that could be settled by textbook definitions I would not have posed it for the Net. The issue is not whether or not topologists have a coherent continuous/discrete distinction but (among other things) whether that distinction (1) corresponds to the A/D Distinction, (2) captures the intuitions, usage and practice of the several disciplines puporting to use the distinction and (3) conforms with physical nature. ----- (8) aweinste@Diamond.BBN.COM (Anders Weinstein) replies on net.ai,net.cog-eng to an earlier iteration about the philosopher Nelson Goodman's formulation: > Well you asked for a "precise" definition! Although Goodman's rigor > may seem daunting, there are really only two main concepts to grasp: > "density", which is familiar to many from mathematics, and > "differentiation". > Goodman mentions that the difference between continuity and density > is immaterial for his purposes, since density is always sufficient to > destroy differentiation (and hence "notationality" and "digitality" as > well). There seems to be some difference of opinion on this matter from the continuity enthusiasts, although they all advocate precision and rigor... > "Differentiation" pertains to our ability to make the necessary > distinctions between elements. There are two sides to the requirement: > "syntactic differentiation" requires that tokens belonging to distinct > characters be at least theoretically discriminable; "semantic > differentiation" requires that objects denoted by non-coextensive > characters be theoretically discriminable as well. > Objects fail to be even theoretically discriminable if they can be > arbitrarily similar and still count as different. Do you mean cases like 2 vs. 1.9999999..., or cases like 2 vs. 2 minus epsilon? They both seem as if they could be either "theoretically discriminable" or "theoretically indiscriminable," depending on the theory. > For example, consider a language consisting of straight marks such > that marks differing in length by even the smallest fraction of an inch > are stipulated to belong to different characters. This language is not > finitely differentiated in Goodman's sense. If, however, we decree > that all marks between 1 and 2 inches long belong to one character, all > marks between 3 and 4 inches long belong to another, all marks between > 5 and 6 inches long belong to another, and so on, then the language > WILL qualify as differentiated. > The upshot of Goodman's requirement is that if a symbol system is to > count as "digital" (or as "notational"), there must be some finite > sized "gaps", however minute, between the distinct elements that need > to be distinguished. > Some examples:... musical notation [vs]... [an unquantized] scale > drawing of a building > To quote Goodman: > "Consider an ordinary watch without a second hand. The hour-hand is > normally used to pick out one of twelve divisions of the half-day. > It speaks notationally [and digitally -- AW]. So does the minute hand > if used only to pick out one of sixty divisions of the hour; but if > the absolute distance of the minute hand beyond the preceding mark is > taken as indicating the absolute time elapsed since that mark was > passed, the symbol system is non-notational. Of course, if we set > some limit -- whether of a half minute or one second or less -- upon > the fineness of judgment so to be made, the scheme here too may > become notational." So apparently it does not matter whether the watch is in fact an "analog" or "digital" watch (according to someone else's definition); according to Goodman's the critical factor is how it's used. > I'm still thinking about your question of how Goodman's distinction > relates to the intuitive notion as employed by engineers or > cognitivists and will reply later. Please be sure to take into consideration the heterogenous sample of replies and rival intuitions this challenge has elicited from these various disciplines. -- Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Fri Nov 7 19:19:08 1986 Date: Fri, 7 Nov 86 19:18:59 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #249 Status: R AIList Digest Wednesday, 5 Nov 1986 Volume 4 : Issue 249 Today's Topics: Philosophy - The Analog/Digital Distinction ---------------------------------------------------------------------- Date: 31 Oct 86 02:45:56 GMT From: husc6!Diamond!aweinste@think.com (Anders Weinstein) Subject: Re: The Analog/Digital Distinction In article <20@mind.UUCP> harnad@mind.UUCP (Stevan Harnad) writes: > I suspect that some people will find Goodman's >considerations a little "dense," not to say hirsute, ... Well you asked for a "precise" definition! Although Goodman's rigor may seem daunting, there are really only two main concepts to grasp: "density", which is familiar to many from mathematics, and "differentiation". >> A scheme is syntactically dense if it provides for infinitely many >> characters so ordered that between each two there is a third. > >I'm no mathematician, but it seems to me that this is not strong >enough for the continuity of the real number line. The rational >numbers are "syntactically dense" according to this definition. But >maybe you don't want real continuity...? Quite right. Goodman mentions that the difference between continuity and density is immaterial for his purposes, since density is always sufficient to destroy differentiation (and hence "notationality" and "digitality" as well). "Differentiation" pertains to our ability to make the necessary distinctions between elements. There are two sides to the requirement: "syntactic differentiation" requires that tokens belonging to distinct characters be at least theoretically discriminable; "semantic differentiation" requires that objects denoted by non-coextensive characters be theoretically discriminable as well. Objects fail to be even theoretically discriminable if they can be arbitrarily similar and still count as different. For example, consider a language consisting of straight marks such that marks differing in length by even the smallest fraction of an inch are stipulated to belong to different characters. This language is not finitely differentiated in Goodman's sense. If, however, we decree that all marks between 1 and 2 inches long belong to one character, all marks between 3 and 4 inches long belong to another, all marks between 5 and 6 inches long belong to another, and so on, then the language WILL qualify as differentiated. The upshot of Goodman's requirement is that if a symbol system is to count as "digital" (or as "notational"), there must be some finite sized "gaps", however minute, between the distinct elements that need to be distinguished. Some examples: A score in musical notation can, if certain conventions are adopted, be regarded as a digital representation, with the score denoting any performance that complies with it. Note that although musical pitches, say, may take on a continuous range of values, once we adopt some conventions about how much variation in pitch is to be tolerated among the compliants of each note, the set of note extensions can become finitely differentiated. A scale drawing of a building, on the other hand, usually functions as an analog representation: any difference in a line's length, however fine, is regarded as denoting a corresponding difference in the building's size. If we decide to interpret the drawing in some "quantized" way, however, then it can be a digital representation. To quote Goodman: Consider an ordinary watch without a second hand. The hour-hand is normally used to pick out one of twelve divisions of the half-day. It speaks notationally [and digitally -- AW]. So does the minute hand if used only to pick out one of sixty divisions of the hour; but if the absolute distance of the minute hand beyond the preceding mark is taken as indicating the absolute time elapsed since that mark was passed, the symbol system is non-notational. Of course, if we set some limit -- whether of a half minute or one second or less -- upon the fineness of judgment so to be made, the scheme here too may become notational. I'm still thinking about your question of how Goodman's distinction relates to the intuitive notion as employed by engineers or cognitivists and will reply later. Anders Weinstein ------------------------------ Date: 26 Oct 86 16:37:53 GMT From: rutgers!princeton!rocksvax!oswego!dl@spam.ISTC.SRI.COM (Doug Lea) Subject: Re: The Analog/Digital Distinction: Soliciting Definitions re: The analog/digital distinction First, I propose a simple ground-rule. Let's assume that the "world" somehow really is "discrete", that is, time, energy, mass, etc., all come in little quanta. Given this, the differences between analog and digital processes seem forced to lie in the nature of representations, algorithms to manipulate them, and the relations of both to actual quantities "out there". I offer a very simple example to illustrate some possibilities. It is intended to be somewhat removed from the sorts of interesting problems encountered in distinguishing analog from digital mental processes. Consider different approaches to determining population growth, given this grossly simplistic model: an initial population, P, a time period in question, T, (expressed in time quanta), and a "growth rate", R, the number of quanta between the times that each member of this (asexual) population gives birth to a new member (supposing that no more than one birth per quantum is possible and no deaths). Approach 1: (digital) Simulate this process with an O(PT) algorithm, repeating T times a scan across each member of the population, determining whether it gave birth, and if so, adding a new member. If the population actually does grow in this fashion, then the result is surely correct, as one might verify by mapping the representation of the P individuals to real individuals at time 0, and again at time T. Several efficiency improvements to this algorithm are, of course, possible. Approach 2: (analog) An alternative method may be constructed by first noting that both population size and time have very simple properties with respect to this process. For purposes of the problem at hand, the difference between the state of having a population of size N and one of size N+1 lies only in the difference between N and N+1. Similarly with time. To develop an algorithm capitalizing on this, construct a nearly equivalent problem in which population states differ only according to the difference between N and N+epsilon, for any epsilon. Now, we know that if epsilon is infinitessimally small, we can exploit the differential and integral calculus to derive an exponential function describing this process, and compute the population value at time T with one swift calculation. Of course, the answer isn't right: we solved a different problem! But it is close, and methods exist to determine just how close this approximation will be in specific instances. We may even be able to apply particular corrections. Approach 3: (digital) Use techniques developed for difference equations and recurrence relations to come up with an exact answer requiring nearly as little calculation as in the analog approach. Approach 4: (digital?) Place P cents in a bank account with compound interest rate corresponding to R, and then see how much money you have at time T. Approach 5: (analog) Build a RLC integrating circuit with the right parameters. Apply input voltage P and measure the output voltage at time T. Approach 6: (analog?) Observe some process with an exponential probability distribution of events. Apply lots of transformations to get an answer. There are probably many other interesting approaches, but I'll leave it there. Morals: 1. The notion of "analogy" or simulation does not do much to distinguish analog from digital processing. Perhaps due to the nature of our physical world, often there do seem to be more and better analog analogies than digital analogies for many problems. 2. Speed of calculation also seems secondary. For example, the calculus allows manipulation of representations involving infinite numbers of states with a single calculation. But some digital methods are fast too. Similarly with the fact that analog methods sometimes allow compact representations (with single numbers and simple well behaved functions representing entire problems). But one could probably match, one-for-one, problems in which analog and digital approaches were superior with respect to these attributes. This all just amounts to acknowledging that the choice between ANY two algorithms ought to take computational efficiency into account. And, of course, the notion of "symbolic" vs. "non-symbolic" processing plays no role here. All of the above approaches were symbolic in one way or another. 3. The notion of approximation seems to be the most helpful one. Again, for example, processing that involves implicit or explicit use of the calculus can ONLY (given the above ground-rule) provide approximations. Most such processing should probably be considered analog. However, the usual conceptualization of approximation in current use doesn't seem good enough. There are many digital "heuristic" algorithms that are labelled as "approximations". (Worse, discrete computational techniques for numerically solving "analytic" problems like integration are also labelled "approximations" in nearly a reverse sense.) For example, the nearest-neighbor heuristic is considered as an approximation algorithm for the travelling salesperson problem. But this seems to be a different sort of approximation than using exponential equations to solve population problems. I'm not at all sure how to go about dealing with such distinctions. Considerations of the robustness and the arbitrary level of precision for approximations in the first sense might be useful, but aren't the whole story: For example, several clearly digital heuristics also have these properties (see, e.g., Karp's travelling saleperson heuristic), but in somewhat different (e.g., probabalistic) contexts. See J. Pearl's "Heuristics" book for related discussions. Doug Lea Computer Science SUNY Oswego Oswego, NY 13126 seismo!rochester!rocksvax!oswego!dl ------------------------------ Date: 4 Nov 86 01:55:22 GMT From: rutgers!husc6!Diamond!aweinste@SPAM.ISTC.SRI.COM (Anders Weinstein) Subject: Re: Analog/Digital Distinction: 8 more replies >Stevan Harnad: > >> Goodman mentions that the difference between continuity and density >> is immaterial for his purposes, since density is always sufficient to >> destroy differentiation (and hence "notationality" and "digitality" as >> well). > >There seems to be some difference of opinion on this matter from the >continuity enthusiasts, although they all advocate precision and rigor... I don't believe there's any major difference here. The respondants who require "continuity" are thinking only in terms of physics, where you don't encounter magnitudes with dense but non-continuous ranges. Goodman deals with other, artificially constructed symbol systems as well. In these we can, by fiat, obtain a scheme that is dense but non-continuous. I think that representation in such a scheme would fit most people's intuitive sense of "analog-icity" if they thought about it. >> Objects fail to be even theoretically discriminable if they can be >> arbitrarily similar and still count as different. > >Do you mean cases like 2 vs. 1.9999999..., or cases like 2 vs. 2 minus epsilon? >They both seem as if they could be either "theoretically >discriminable" or "theoretically indiscriminable," depending on the >theory. I'm not sure what you mean here. I don't see how a length of 2 inches would count as "theoretically discriminable" from a length of 1.999... inches; nor is a length of 2 inches theoretically discriminable from a length of 2 minus epision inches if epsilon is allowed to be arbitrarily small. On the other hand, a length of 2 inches IS theoretically discriminable from a length of 1.9 inches. In his examples, Goodman rules out cases where no measurement of any finite degree of precision would be sufficient to make the requisite distinctions. >> "Consider an ordinary watch without a second hand. The hour-hand is >> normally used to pick out one of twelve divisions of the half-day. >> It speaks notationally [and digitally -- AW]. So does the minute hand >> if used only to pick out one of sixty divisions of the hour; but if >> the absolute distance of the minute hand beyond the preceding mark is >> taken as indicating the absolute time elapsed since that mark was >> passed, the symbol system is non-notational. Of course, if we set >> some limit -- whether of a half minute or one second or less -- upon >> the fineness of judgment so to be made, the scheme here too may >> become notational." > >So apparently it does not matter whether the watch is in fact an >"analog" or "digital" watch (according to someone else's definition); >according to Goodman's the critical factor is how it's used. Right. Remember, Goodman is not talking about whether this is what an engineer would class as an analog or digital WATCH (ie. in its internal workings); he's ONLY talking about the symbol system used to represent the time to the viewer. And he's totally relativistic here -- whether the representation is analog or digital depends entirely on how it is to be read. ------------------------------ Date: Tue, 4 Nov 86 17:10:39 pst From: ladkin@kestrel.ARPA (Peter Ladkin) Subject: analog/digital distinction Here's a quick shot at an A/D distinction. The problem with the rationals was that the ordering and the operations are easily translatable into computations on the natural numbers. So, the proposal is: DIGITAL: computations on a structure S that is recursively isomorphic to a definable fragment of Peano Arithmetic. ANALOG: computations on a dense structure that is not recursively isomorphic to a definable fragment of Peano Arithmetic. Note there can be computations which are neither analog nor digital according to this definition. The rationale for this choice depends on two considerations. (1) One must not be able to transform one kind of computation into the other, which can be done only if there is a machine (aka recursive function) that can do it. (2) The distinction must not collapse in the face of the possibility that physics will tell us the world is fundamentally discrete (or fundamentally continuous), since if Gerald Holton is to be believed, physical science has been wavering between one and the other for thousands of years. So the discrete/continuous nature of nature can be regarded as a metaphysical issue, and we want to finesse this in our definition to make it physically realistic. I chose Peano Arithmetic as the base structure because it is intuitively discrete, and all the digital structures that have been proposed fit the criterion that they can be recursively mapped into simple discrete arithmetic. The density-of-values criterion for analog computation seems intuitively plausible, and if one wants to make the distinction between analog and digital into a feature of the world, not merely of the representation chosen, one needs to assure consideration (1) above. If quantum physics ultimately tells us that the world is discrete, there is no reason to assume that the discreteness in the world will provide us with recursive functions mapping that discreteness into the natural numbers, so analog computations will survive that discovery. Peter Ladkin ladkin@kestrel.arpa ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Sun Nov 9 02:21:17 1986 Date: Sun, 9 Nov 86 02:21:07 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #250 Status: R AIList Digest Thursday, 6 Nov 1986 Volume 4 : Issue 250 Today's Topics: Philosophy - The Analog/Digital Distinction & Information ---------------------------------------------------------------------- Date: 29 Oct 86 16:29:10 GMT From: rutgers!princeton!mind!harnad@titan.arc.nasa.gov (Stevan Harnad) Subject: The A/D Distinction: 5 More Replies [This message actually fits into the middle of the sequence I sent yesterday. Sorry for the reversal. -- KIL] Here are 5 more replies I've received on the A/D distinction. I'll respond in a later module. [Meantime, could someone post this to sci.electronics, to which I have no access, please?] ------ (1) Message-Id: <8610271622.11564@ur-seneca.arpa> In-Reply-To: <13@mind.UUCP> U of Rochester, CS Dept, Rochester, NY ken@rochester.arpa CS Dept., U. of Roch., NY 14627. Mon, 27 Oct 86 11:22:10 -0500 I think the distinction is simply this: digital deals with a finite set of discrete {voltage, current, whatever} levels, while analog deals with a *potentially* infinite set of levels. Now I know you are going to say that analog is discrete at the electron noise level but the circuits are built on the assumption that the spectrum is continuous. This leads to different mathematical analyses. Sort of like infinite memory Turing machines, we don't have them but we program computers as if they had infinite memory and in practice as long as we don't run out, it's ok. So as long as we don't notice the noise in analog, it serves. -------- (2) Tue, 28 Oct 86 20:56:36 est cuuxb!mwm AT&T-IS, Software Support, Lisle IL In article <7@mind.UUCP> you write: >The ground-rules are these: Try to propose a clear and >objective definition of the analog/digital distinction that is not >arbitrary, relative, a matter of degree, or loses in the limit the >intuitive distinction it was intended to capture. > >One prima facie non-starter: "continuous" vs. "discrete" physical >processes. > >Stevan Harnad (princeton!mind!harnad) Analog and digital are two ways of *representing* information. A computer can be said to be analog or digital (or both!) depending upon how the information is represented within the machine, and particularly, how the information is represented when actual computation takes place. Digital is essentially a subset of analog, where the range of properties used to represent information is grouped into a finite set of values. For example, the classic TTL digital model uses electrical voltage to represent values, and is grouped into the following: above +5 volts -- not used +2..+5 volts (approx) -- a binary 1 0..+2 volts (approx) -- a binary 0 negatvie voltage -- not used. Important to distinguish here is the grouping of the essentially infinite possiblities of voltage into a finite set of values. A system that used 4 voltage ranges to represent a base 4 number system would still be digital. Note that this means that it takes several voltages to represent an arbitrarily precise number Analog, on the other hand, refers to using a property to directly represent an infinite range of values with a different infinite range of values: for example representing the number 15 with 15 volts, and the number 100 with 100 volts. Note that this means it takes 1 voltage to represent an arbitrarily precise number. This is my pot-shot at defining analog/digital and how they relate, and how they are used in most systems i am familiar with. I think these make reasonably clear what it is that "analog to digital" converters (and "digital to analog") do. This is why slide-rules are considered analog, you are using distance rather than voltage, but you can interpret a distance as precisely as you want. An abacus, on the otherhand also uses distance, but where a disk is means either one thing or another, and it takes lots of disks to represent a number. An abacus then, is digital. Marc Mengel ...!ihnp4!cuuxb!mwm -------- (3) Thu, 23 Oct 86 13:10:47 pdt Message-Id: <8610232010.AA18462@bcsaic.LOCAL> Try this: (An) analog is a (partial) DUPLICATE (or abstraction) of some material thing or some process, which contains (it is hoped) the significant characteristics and properties of the original. An analog is driven by situations and events outside itself, and its usefulness is that the analog may be observed and, via induction, the original understood. A digital device or method operates on symbols, rather than physical (or other) reality. Analog computers may operate on (real) voltages and electron flow, while digital computers operate on symbols and their logical interrelationships. Digital operations are formal; that is they treat form rather than content, and are therefore always deductive, while the behavior of real things and their analogs is not. (Heresy follows). It is one of my (unpopular) assertions that the central nervous system of living organisms (including myself) is best understood as an analog of "reality"; that most interesting behavior such as induction and the detection of similarity (analogy and metaphor) cannot be accomplished with only symbolic, and therefore deductive, methods. -------- (4) Mon, 27 Oct 86 16:04:36 mst lanl!a.LANL.ARPA!crs (Charlie Sorsby) Message-Id: <8610272304.AA25429@a.ARPA> References: <7@mind.UUCP> <45900003@orstcs.UUCP>, <13@mind.UUCP> Stevan, I've been more or less following your query and the resulting articles. It seems to me that the terms as they are *usually* used today are rather bastardized. Don't you think that when the two terms originated they referred to two ways of "computing" and *not* to kinds of circuits at all. The analog simulator (or, more popularly, analog computer) "computed" by analogy. And, old timers may recall, they weren't all electronic or even electrical. I vaguely recall reading about an analog simultaneous linear-equation solver that comprised plates (rectangular, I think), cables and pulleys. Digital computers (truly so) on the other hand computed with *digits* (i.e. numbers). Of course there was (is) analogy involved here too but that was a "higher-order term" in the view and was conveniently ignored as higher order terms often are. In the course of time, the term analog came to be used for those electronic circuits *like* those used in analog simulators (i.e. circuits that work with continuous quantities). And, of course, digital came to refer to those circuits *like* those used in digital computers (i.e. those which work with discrete or quantized quantities. Whether a quantity is continuous or discrete depends on such things as the attribute considered to say nothing of the person doing the considering, hence the vagueness of definition and usage of the terms. This vagueness seems to have worsened with the passage of time. Best regards, Charlie Sorsby ...!{cmcl2,ihnp4,...}!lanl!crs crs@lanl.arpa -------- (5) Message-Id: <8610280022.AA16966@mitre.ARPA> Organization: The MITRE Corp., Washington, D.C. sundt@mitre.ARPA Date: Mon, 27 Oct 86 19:22:21 -0500 Having read your messages for the last few months, I couldn't help but take a stab on this issue. Coming from a heavily theoretical undergraduate physics background, it seems obvious that the ONLY distinction between the analog and digital representation is the enumerability of the relationships under the given representation. First of all, the form of digital representation must be split into two categories, that of a finite representation, and that of a countably infinite representation. Turing machines assume a countably infinite representation, whereas any physically realizable digital computer must inherently assume a finite digital representation (be it ever so large). Thus, we have three distinctions to make: 1) Analog / Finite Digital 2) Countably-Infinite Digital / Finite Digital 3) Analog / Countably-Infinite Digital Second, there must be some predicate O(a,b) defined over all the a and b in the representation such that the predicate O(a,b) yields only one of a finite set of symbols, S(i) (e.g. "True/False"). If such a predicate does not exist, then the representation is arguably ambiguous and the symbols are "meaningless". An example of an O(a,b) is the equality predicate over the reals, integers, etc. Looking at all the (a,b) pairs that map the O(a,b) predicate into the individual S(i), note that the following is true: ANALOG REPRESENTATION: the (a,b) pairs cannot be enumerated for ALL S(i). COUNTABLY-INFINITE DIGITAL REPRESENTATION: the (a,b) pairs cannot be enumerated for ALL S(i). FINITE DIGITAL REPRESENTATION: all the (a,b) pairs for all the S(i) CAN be enumerated. This distinguishes the finite digital representation from the other two representations. I believe this is the distinction you were asking about. The distinction between the analog representation and the countably-infinite digital representation is harder to identify. I sense it would require the definition of a mapping M(a,b) onto the representation itself, and the study of how this mapping relates to the O(a,b) predicate. That is, is there some relationship between O(?,?), M(?,?) and the (a,b) that is analgous to divisibility in Z and R. How this would be formulated escapes me. On your other-minds problem: [see "Searle, Turing, Categories, Symbols"] I think the issue here is related to the above classification. In particular, I think the point to be made is that we can characterize when something is NOT intelligent, but are unable to define when it is. A less controversial issue would be to "Define chaos". Any attempt to do so would give it a fixed structure, and therefore order. Thus, we can only define chaos in terms of what it isn't, i.e. "Chaos is anything that cannot be categorized." Thus, it is the quality that is lost when a signal is digitized to either a finite or an countably-infinite digital representation. Analog representations would not suffer this loss of chaos. Carrying this thought back to "intelligence," intelligence is the quality that is lost when the behavior is categorized among a set of values. Thus, to detect intelligence, you must use analog representations ( and meta-representations). And I am forced to conclude that the Turing test must always be inadequate in assessing intelligence, and that you need to be an intelligent being to *know* an intelligent being when you see one!!! Of course, there is much error in categorizations like this, so in the *real* world, a countably-infinite digital representation might be *O.K.*. I wholy agree with your arguement for a basing of symbols on observables, and would also argue that semantic content is purely a result of a rich syntactic structure with only a few primitive predicates, such as set relations, ordering relations, etc. Thinking about it further, I would argue, in view of what I just said, that people are by construction only "faking" intelligence, and that we have achieved a complexity whereby we can percieve *some* of the chaos left by our crude categorizations (perhaps through multiple categorizations of the same phenomena), and that this perception itself gives us the appearance of intelligence. Our perceptions reveal only the tip of the chaotic iceberg, however, by definition. To have true intelligence would require the perception of *ALL* the chaos. I hope you found this entertaining, and am anxious to hear your response. Mitch Sundt The MITRE Corp. sundt@mitre.arpa ------------------------------ Date: 3 Nov 86 23:40:28 GMT From: ladkin@kestrel.ARPA (Peter Ladkin) Subject: Re: The Analog/Digital Distinction (weinstein quoting goodman) > > A scheme is syntactically dense if it provides for infinitely many > > characters so ordered that between each two there is a third. (harnad) > I'm no mathematician, but it seems to me that this is not strong > enough for the continuity of the real number line. The rational > numbers are "syntactically dense" according to this definition. Correct. There is no first-order way of defining the real number line without introducing something like countably infinite sequences and limits as primitives. Moreover, if this is done in a countable language, you are guaranteed that there is a countable model (if the definition isn't contradictory). Since the real line isn't countable, the definition cannot ensure you get the REAL reals. Weinstein wants to identify *analog* with *syntactically dense* plus some other conditions. Harnad observes that the rationals fit the notion of syntactic density. The rationals are, up to isomorphism, the only countable, dense, linear order without endpoints. So any syntactically dense scheme fitting this description is (isomorphic to) the rationals, or a subinterval of the rationals (if left-closed, right-closed, or both-closed at the ends). One consequence is that one could define such an *analog* system from a *digital* one by the following method: Use the well-known way of defining the rationals from the integers - rationals are pairs (a,b) of integers, and (a,b) is *equivalent* to (c,d) iff a.d = b.c. The *equivalence* classes are just the rationals, and they are semantically dense under the ordering (a,b) < (c,d) iff there is (f,g) such that f,g have the same sign and (a,b) + (f,g) = (c,d) where (a,b) + (c,d) = (ad + bc, bd), and the + is factored through the equivalence. We may be committed to this kind of phenomenon, since every plausible suggested definition must have a countable model, unless we include principles about non-countable sets that are independent of set theory. And I conjecture that every suggestion with a countable model is going to be straightforwardly obtainable from the integers, as the above example was. Peter Ladkin ladkin@kestrel.arpa ------------------------------ Date: 3 Nov 86 23:47:34 GMT From: ladkin@kestrel.ARPA (Peter Ladkin) Subject: Re: The Analog/Digital Distinction In article <1701@Diamond.BBN.COM>, aweinste@Diamond.BBN.COM (Anders Weinstein) writes: > The upshot of Goodman's requirement is that if a symbol system is to count as > "digital" (or as "notational"), there must be some finite sized "gaps", > however minute, between the distinct elements that need to be distinguished. I'm not sure you want this definition of the distinction. There are *finite-sized gaps, however minute* between rational numbers, and if we use the pairs-of-integers representation to represent the syntactically dense scheme, (which must be isomorphic to some subrange of the rationals if countable) we may use the integers and their gaps to distinguish the gaps in the syntactically dense scheme, in a quantifier-free manner. Thus syntactically dense schemes would count as *digital*, too. Peter Ladkin ladkin@kestrel.arpa ------------------------------ Date: 4 Nov 86 19:03:09 GMT From: nsc!amdahl!apple!turk@hplabs.hp.com (Ken "Turk" Turkowski) Subject: Re: Analog/Digital Distinction In article <116@mind.UUCP> harnad@mind.UUCP (Stevan Harnad) writes: >(2) winnie!brad (Brad Garton) writes: >> ... When I consider the digitized versions of analog >> signals we deal with over here , it seems that we >> approximate more and more closely the analog signal with the >> digital one as we increase the sampling rate. There is a difference between sampled signals and digital signals. A digital signals is not only sampled, but is also quantized. One can have an analog sampled signal, as with CCD filters. As a practical consideration, all analog signals are band-limited. By the Sampling Theorem, there is a sampling rate at which a bandlimited signal can be perfectly reconstructed. *Increasing the sampling rate beyond this "Nyquist rate" cannot result in higher fidelity*. What can affect the fidelity, however, is the quantization of the samples: the more bits used to represent each sample, the more accurately the signal is represented. This brings us to the subject of Signal Theory. A particular class of signal that is both time- and band-limited (all real-world signals) can be represented by a linear combination of a finite number of basis functions. This is related to the dimensionality of the signal, which is approximately 2WT, where W is the bandwidth of the signal, and T is the duration of the signal. >> ... This process reminds >> me of Mandelbrot's original "How Long is the Coastline of Britain" >> article dealing with fractals. Perhaps "analog" could be thought >> of as the outer limit of some fractal set, with various "digital" >> representations being inner cutoffs. Fractals have a 1/f frequency distribution, and hence are not band-limited. >> In article <105@mind.UUCP> harnad@mind.UUCP (Stevan Harnad) writes: >> I'm not convinced. Common ways of transmitting analog signals all >> *do* lose at least some of the signal, irretrievably... Let's not forget noise. It is impossible to keep noise out of analog channels and signal processing, but it can be removed in digital channels and can be controlled (roundoff errors) in digital signal processing. >> ... Losses of information in processing analog signals tend to >> be worse, and for an analog transformation to be exactly invertible, it >> *must* preserve all the information in its input. Including the exclusion of noise. Once noise is introduced, the signal cannot be exactly inverted. -- Ken Turkowski @ Apple Computer, Inc., Cupertino, CA UUCP: {sun,nsc}!apple!turk CSNET: turk@Apple.CSNET ARPA: turk%Apple@csnet-relay.ARPA ------------------------------ Date: Wed 5 Nov 86 21:03:42-PST From: Ken Laws Subject: Information in Signals From: nsc!amdahl!apple!turk@hplabs.hp.com (Ken "Turk" Turkowski) Message-Id: <267@apple.UUCP> *Increasing the sampling rate beyond this "Nyquist rate" cannot result in higher fidelity*. >>... Losses of information in processing analog signals tend to >>be worse, and for an analog transformation to be exactly invertible, it >>*must* preserve all the information in its input. Including the exclusion of noise. Once noise is introduced, the signal cannot be exactly inverted. To pick a couple of nits: Sampling at the Nyquist rate preserves information, but only if the proper interpolation function is used to reconstruct the continuous signal. Often this function is nonphysical in the sense that it extends infinitely far in each temporal direction and contains negative coefficients that are difficult to implement in some types of analog hardware (e.g., incoherent optics). One of the reasons for going to digital processing is that [approximate] sinc or Bessel functions are easier to deal with in the digital domain. If a sampled signal is simply run through the handiest speaker system or other nonoptimal reconstruction, sampling at a higher rate may indeed increase fidelity. The other two quotes are talking about two different things. No transformation (analog or digital) is invertible if it loses information, but adding noise to a signal may or may not degrade its information content. An analog signal can be just as redundant as any coded digital signal -- in fact, most digital "signals" are actually continuous encodings of discrete sequences. To talk about invertibility one must define the information in a signal -- which, unfortunately, depends on the observer's knowledge as much as it does on the degrees of freedom or joint probability distribution of the signal elements. Even "degree of freedom" and "probability" are not well defined, so that our theories are ultimately grounded in faith and custom. Fortunately the real world is kind: our theories tend to be useful and even robust despite the lack of firm foundations. Philosophers may demonstrate that engineers are building houses of cards on shifting sands, but the engineers will build as long as their houses continue to stand. -- Ken Laws ------------------------------ Date: Wed, 5 Nov 1986 16:00 EST From: MINSKY%OZ.AI.MIT.EDU@XX.LCS.MIT.EDU Subject: AIList Digest V4 #248 With all due respect, I wonder if the digital-analog discussion could be tabled soon. I myself do not consider it useful to catalog the dispositions of many different persons' use of a word; in any case the thing has simply gone past the bounds of 1200 baud communication. Please. On to some substance. ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Sun Nov 9 02:20:13 1986 Date: Sun, 9 Nov 86 02:20:08 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #251 Status: R AIList Digest Thursday, 6 Nov 1986 Volume 4 : Issue 251 Today's Topics: Queries - Franz Object-Oriented Packages & Sentient-Computer Novels & Simulating a Neural Network, Application - Snooker-Playing Robots, Ethics - Moral Responsibility, Seminars - Planning Simultaneous Actions (UPenn) & Scientific Discovery (CMU) & Machine Inductive Inference (CMU) & Case-Based Learning System (Rutgers) ---------------------------------------------------------------------- Date: Wed, 5 Nov 86 13:08:28 EST From: weltyc%cieunix@CSV.RPI.EDU (Christopher A. Welty) Subject: Looking for Franz OO packages I am looking for information on Object Oriented extensions to Franz Lisp. I know that someone (U of Maryland?) came out with a flavors package for Franz, if someone can point me in the right direction there it would be appreciated, as well as any info on other packages... ------------------------------ Date: 5 Nov 86 23:45:05 GMT From: gknight@ngp.utexas.edu (Gary Knight) Subject: Canonical list of sentient computer novels I am trying to compile a canonical list of SF *novels* dealing with (1) sentient computers, and (2) human mental access to computers or computer networks. Examples of the two categories (and my particular favorites as well) are: A) SENTIENT COMPUTERS The Adolescence of P-1, by Thomas J. Ryan Valentina: Soul in Sapphire, by Joseph H. Delaney and Marc Stiegler Cybernetic Samurai, by (I forget) Coils, by Roger Zelazny B) HUMAN ACCESS True Names, by Vernor Vinge Neuromancer and Count Zero, by William Gibson I'm not sure how this is done, but my thought is for all of you sf-fans out there to send me e-mail lists of such novels (separate, by category A and B), and I'll compile and post the ultimate canonical version. I've heard that this exercise was undertaken a year or so ago, but I don't have access to that list and besides I'd like to get fresh input anyway (and recent qualifying books). So let me hear from you . . . . Gary -- Gary Knight, 3604 Pinnacle Road, Austin, TX 78746 (512/328-2480). Biopsychology Program, Univ. of Texas at Austin. "There is nothing better in life than to have a goal and be working toward it." -- Goethe. ------------------------------ Date: 30 Oct 86 15:20:24 GMT From: ihnp4!inuxc!iuvax!cdaf@ucbvax.Berkeley.EDU (Charles Daffinger) Subject: Re: simulating a neural network In article <151@uwslh.UUCP> lishka@uwslh.UUCP [Chris Lishka] writes: > >... > Apparently Bell Labs (I think) has been experimenting with neural >network-like chips, with resistors replacing bytes (I guess). They started >out with about 22 'neurons' and have gotten up to 256 or 512 (can't >remember which) 'neurons' on one chip now. Apparently these 'neurons' are >supposed to run much faster than human neurons... ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ What bothers me is that the performance is rated upon speed. Unlike the typical syncronous digital computer, neuronal networks are asyncronous, communicating via a temporal discharge of 'spikes' through axons which vary considerably in length, as well as speed, and exploit the use of SLOW signals just as they do those of FAST signals. (look at the neral mechanism for a reflex, or for that of focusing the eye, as an example). I am curious as to how much of the essence of their namesakes was really captured in these 'neurons'? -charles -- ... You raise the blade, you make the change, you re-arrange me til I'm sane... Pink Floyd ------------------------------ Date: 3 Nov 1986 13:49:29 GMT From: Icarus Sparry Subject: Snooker playing robots This is being posted on behalf of another member of staff, who is not able to get through the UCL gateway ------ Newsgroups: mod.ai Subject: Re: Robot Snooker-player Summary: Expires: References: <861020-061334-1337@Xerox> Sender: Reply-To: cc_dgdc@ux63.bath.ac.uk (Clark) Followup-To: Distribution: Organization: University of Bath, England Keywords: I believe you will find the robot snooker player at Bristol University, England. I too saw a local tv news program about it last year. I think the AI group is in one of the Engineering Departments. Doug Clark Bath University ---------- Icarus Mr I. W. J. Sparry Phone 0225 826826 x 5983 Head of Microcomputer Unit Telex 449097 University of Bath e-mail: Claverton Down cc_is@UK.AC.BATH.UX63 Bath BA2 7AY !mcvax!ukc!hlh!bath63!cc_is England cc_is%ux63.bath.ac.uk@ucl-cs.arpa ------------------------------ Date: Wed, 5 Nov 86 12:25:26 est From: Randy Goebel LPAIG Subject: Re: moral responsibility Patrick Hayes writes > ...Weizenbaum has made a successful career by > systematically attacking AI research on the grounds that it is somehow > immoral, and finding a large and willing audience. Weizenbaum does, indeed and unfortunately, attract a large, willing and naive audience. For some reason, there seems to be a large not-quite- computer-literate population that wants to believe that AI is potentially dangerous to ``real'' intelligence. But it is not completely fair to conclude that Weizenbaum believes AI to be immoral; it is correct for Patrick to qualify his conclusion as ``somehow'' immoral. Weizenbaum acknowledges the general concept of intelligence, with both human and artificial kinds as manifestations. He even prefers the methodology of the artificial kind, especially when it relieves us from experiments on, say, the visual cortex of cats. Weizenbaum does claim that certain aspects of AI are immoral but, as the helicopter example illustrates, his judgment is not exclusive to AI. As AI encroaches most closely to those things Weizenbaum values (e.g., human dignity, human life, human emotions), it is natural for him to speak about the potential dangers that AI poses. I suspect that, if Weizenbaum were a nuclear physicist instead of a computer scientist, he would focus more attention on the immorality of fission and fusion. It is Weizenbaum's own principles of morality that determine the judgements. He acknowledges that, and places his prinicples in the public forum every time he speaks. ------------------------------ Date: Mon, 3 Nov 86 14:27 EST From: Tim Finin Subject: Seminar - Planning Simultaneous Actions (UPenn) Computer and Information Science Colloquium University of Pennsylvania 3-4:30 pm Thursday, November 6, 1986 Room 216 - Moore School PLANNING SIMULTANEOUS ACTIONS IN TEMPORALLY RICH WORLDS Professor James Allen Department of Computer Science University of Rochester This talk describes work done with Richard Pelavin over the last few years. We have developed a formal logic of action that allows us to represent knowledge and reason about the interactions between events that occur simultaneously or overlap in time. This includes interactions between two (or more) actions that a single agent might perform simultaneously, as well as interactions between an agent's actions and events occuring in the external world. The logic is built upon an interval-based temporal logic extended with modal operators similar to temporal necessity and a counterfactual operator. Using this formalism, we can represent a wide range of possible ways in which actions may interact. ------------------------------ Date: 4 Nov 86 15:44:08 EST From: Steven.Minton@k.cs.cmu.edu Subject: Seminar - Scientific Discovery (CMU) As usual, 3:15 in 7220. This week's speaker is Deepak Kulkarni. Title: Processes of scientific discovery: Strategy of Experimentation KEKADA is a program that models some strategies of experimentation which scientists use in their research. When augmented with appropriate background knowledge, it can simulate in detail Krebs' course of discovery of urea synthesis. Williamson's discovery of alcohol-structure is another discovery it can simulate. I would like to discuss the general mechanisms used in the system and some half-baked ideas about further work on the system. ----- Deepak told me that he's very interested in getting feedback on some of his ideas for further work. I'm hoping that we'll have a lively feedback session. - Steve ------------------------------ Date: 27 Oct 86 14:20:41 EST From: Lydia.Defilippo@cad.cs.cmu.edu Subject: Seminar - Machine Inductive Inference (CMU) Dates: 3-Nov-86 Time: 4:00 Cboards: general Place: 223d Porter Hall Type: Philosophy Colloquium Duration: one hour Who: Scott Weinstein, University of Pennsylvania Topic: Some Recent Results in the Theory of Machine Inductive Inference Host: Dan Hausman The talk will describe recent research by Dan Osherson, Mike Stob and myself on a variety of topics of epistemological interest in the theory of machine inductive inference. The topics covered will include limitations on mechanical realizations of Bayesian inference methods, the synthesis of inference machines from descriptions of the problem domains for which they are intended and the identification of relational structures. ------------------------------ Date: 29 Oct 86 22:57:40 EST From: Tom Fawcett Subject: Seminar - Case-Based Learning System (Rutgers) TITLE: Memory Access Techniques for a Case-based Learning System SPEAKER: Wendy Lehnert DATE: Monday, November 3 LOCATION: Princeton University, Green Hall, Langfeld Lounge TIME: 12:00 - 1:00 p.m. Abstract Traditionally, symbolic processing techniques in artificial intelligence have addressed "high-level" cognitive tasks like expert reasoning, natural language processing, and knowledge acquisition. At the same time, a separate paradigm of connectionist techniques has addressed "low-level" perceptual problems like word recognition, stereoscopic vision and speech recognition. While symbolic computation models are frequently characterized as brittle, difficult to extend, and exceedingly fragile, many connectionist models exhibit graceful degradation and natural methodologies for system expansion. In this talk, we will look at how connectionist techniques might be useful as a strategy for indexing symbolic memory. Our discussion will focus on two seemingly unrelated tasks: word pronunciation and narrative summarization. We will endeavor to show how both problems can be approached with similar strategies for indexing memory and resolving competing indices. ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Sun Nov 9 02:20:36 1986 Date: Sun, 9 Nov 86 02:20:29 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #252 Status: R AIList Digest Thursday, 6 Nov 1986 Volume 4 : Issue 252 Today's Topics: Funding - NSF Knowledge and Database Systems Awards ---------------------------------------------------------------------- Date: Fri 31 Oct 86 11:31:36-CST From: ICS.DEKEN@R20.UTEXAS.EDU Subject: Knowledge and Database Systems Awards - NSF Fiscal Year 1986 Research Projects Funded by the Information Science Program (now Knowledge and Database Systems Program) A complete listing of these awards, including short descriptive abstracts of the research is available by writing to: Joseph Deken, Director Knowledge and Database Systems Program National Science Foundation 1800 G Street NW Washington, DC 20550 IST-8504726 $42,196 - 12 mos. James F. Allen University of Rochester Plan-Based Approaches to Extended Dialogues - - - BNS-8518675 $40,000 - 12 mos. James A. Anderson Brown University Cognitive Applications of Matrix Memory Models - - - IST-8511531 $19,097 - 12 mos. Robert Berwick Massachusetts Institute of Technology Learnability and Parsability - - - DCR-8603231 $27,000 - 12 mos. Alan W. Biermann Duke University Dialog Processing for Voice Interactive Problem Solving - - - IST-8612177 $9,747 - 12 mos. Jeffrey Bonar University of Pittsburgh Partial Support for Third International Conference on Artificial Intelligence and Education, Pittsburgh, PA, May 1987 - - - IST-8604923 $64,660 - 12 mos. Alan H. Borning University of Washington Automatic Generation of Interactive Displays - - - IST-8643739 $121,074 - 12 mos. Bruce C. Buchanan Stanford University Information Structure and Use in Knowledge-Based Expert Systems - - - IST-8607303 $10,000 - 12 mos. Kathleen M. Carley Carnegie-Mellon University Knowledge Acquisition as a Social Phenomenon - - - IST-8515005 $75,900 - 12 mos. Eugene Charniak Brown Unive A Single-Semantic-Process Theory of Parsing - - - IST-8644629 $60,449 - 12 mos. Eugene Charniak Brown University An Approach to Abductive Inference in Artificial Intelligence Systems - - - IST-8608362 $172,9212 - 12 mos. Richard Cullingford Georgia Institute of Technology Robust Interaction and Natural Problem-Solving in Advice-Giving Systems - - - IST-8506706 $60,942 - 12 mos. Donald Dearholt New Mexico State University Properties of Networks Derived from Proximities - - - IST-8518706 $37,243 - 12 mos. Andre de Korvin Indiana University Foundation Modeling Goal Uncertainty and Goal Shaping in a Generalized Information System - - - IST-8609441 $20,205 - 6 mos. Michael L. Dertouzos Massachusetts Inst. Tech. Conference on Cellular Automata: Parallel Information Processing for Mathematics and Science - - - IST-8519926 $73,460 - 12 mos. Thomas G. Dietterich Oregon State University Learning by Experimentation - - - IST-8519924 $43,489 - 12 mos. John W. DuBois University of California at Los Angeles Information Transfer Constraints and Strategies in Natural Language Communication - - - DCR-8602385 $25,000 - 12 mos. Wayne Dyksen and Mikhail Atallah Purdue University High Level Systems for Scientific Computing - - - IST-8609123 $93,156 - 24 mos. Andrew U. Frank University of Maine at Orono A Formal Model for Representation and Manipulation of Spatial Subdivisions in Information Systems - - - IST-8611673 $18,000 - 3 mos. Thomas Gay University of Connecticut Health Center Travel Grant: U.S. - U.S.S.R. Symposium on Information Coding and Transmission in Biological Systems, October 3-13, 1986 - - - IST-8512419 $70,867 - 12 mos. Richard Granger University of California, Irvine Unification of Lexical, Syntactic, and Pragmatic Inference in Understanding - - - IST-8509860 $73,479 - 12 mos. Robert M. Gray Stanford University The Application of Information Theory to Pattern Recognition and the Design of Decision Tree Classifiers - - - IST-8603943 $134,694 - 18 mos. Max Henrion Carnegie Mellon University A Comparison of Methods for Representing Uncertainty in Expert Systems - - - DCR-8608311 $45,000 - 12 mos. Lawrence J. Henschen Northwestern University Logic and Databases - - - IST-8645349 $153,289 - 12 mos. Richard J. Herrnstein Harvard University A Comparative Approach to Natural and Artificial Visual Information Processing - - - IST-8520359 $70,735 - 12 mos. Geoffrey Hinton Carnegie-Mellon University Search Methods for Massively Parallel Networks - - - IST-8511541 $69,815 - 12 mos. Richard B. Hull University of Southern California Investigation of Practical and Theoretical Aspects of Semantic Database Models - - - IST-8643740 $98,507 - 12 mos. Ray Jackendoff and Jane Grimshaw Brandeis University Syntactic and Semantic Information in a Natural Language Lexicon - - - IST-8512108 $99,394 - 12 mos. Hans Kamp University of Texas at Austin Logic Representation of Attitudes for Computer Natural Language Understanding - - - IST-8644864 $35,721 - 12 mos. Abraham Kandel Florida State University Analysis and Modeling of Imprecise Information in Uncertain Environments - - - IST-8542811 $65,414 - 12 mos. R.L. Kashyap Purdue University Research on Inference Procedures with Uncertainty - - - IST-8644676 $74,752 - 12 mos. George J. Klir State University of New York at Binghamton Possibilistic Information: Theory and Applicability - - - IST-8552925 $54,250 - 12 mos. Richard E. Korf University of California at Los Angeles Presidential Young Investigator Award : Machine Learning - - - IST-8518307 $15,750 - 12 mos. Donald H. Kraft Louisiana State University Travel to the ACM Conference on Research and Development in Information Retrieval: Pisa, Italy; September 8-10, l986 - - - DCR-8602665 $45,720 - 12 mos. Benjamin J. Kuipers University of Texas at Austin Knowledge Representations for Expert Causal Models - - - RII-8600412 $10,000 - 12 mos. Jill H. Larkin University of California at Berkeley Developing the Instructional Power of Modern Personal Computing - - - IST-8600412 $10,000 - 12 mos. Wendy G. Lehnert University of Massachusetts at Amherst Presidential Young Investigator Award: Natural Language Computing Systems - - - IST-8603697 $5,000 - 12 mos. Michael E. Lesk Bell Communications Research Workshop on Document Generation Principles - - - IST-8602765 $76078 - 12 mos. R. Duncan Luce Harvard University Measurement: Axiomatic and Meaningfulness Studies - - - IST-8444028 $62,500 - 12 mos. David Maier Oregon Graduate Center Presidential Young Investigator Award: Foundations of Knowledge Management Systems - - - IST-8604977 $46,956 - 12 mos. David Maier Oregon Graduate Center Automatic Generation of Interactive Displays - - - IST-8642813 $25,500 - 12 mos. Gerald S. Malecki Office of Naval Research Committee on Human Factors - - - IST-8606187 $19,650 - 12 mos. James L. McClelland Carnegie-Mellon University Workshop of Parallel Distributed Processing in Information and Cognitive Research (Washington, D.C. ; February 28 - March 1, 1986) - - - IST-8451438 $37,500 - 12 mos. Kathleen R. McKeown Columbia University Presidential Young Investigator Award: Natural Language Interfaces - - - IST-8520217 $115,220 - 12 mos. Douglas P. Metzler University of Pittsburgh An Expert System Approach to Syntactic Parsing and Information Retrieval - - - IST-8512736 $137,503 - 24 mos. David Mumford Harvard University The Parsing of Images - - - IST-8604282 $2,447 - 12 mos. Kent Norman University of Maryland at College Park Developing an Effective User Evaluation Questionnaire for Interactive Systems - - - IST-8645347 $80,492 - 12 mos. Donald E. Nute University of Georgia Discourse Representation for Natural Language Processing - - - IST-8645348 $49,211 - 12 mos. Donald E. Nute University of Georgia Hypothetical Reasoning and Logic Programming - - - IST-8642477 $86,969 - 12 mos. Robert N. Oddy Syracuse University Representations for Anomalous States of Knowledge in Information Retrieval - - - IST-8609201 $44,905 - 12 mos. Daniel Osherson Syracuse University A Computational Approach to Decision-Making - - - IST-8544976 $197,055 - 12 mos. Charles Parsons and Isaac Levi Columbia University The Structure of Information in Science: Fact Formulas and Discussion Structures in Related Subsciences - - - IST-8642841 $12,000 - 12 mos. William J. Rapaport State University of New York - System Office Logical Foundations for Belief Representation - - - IST-8644984 $62,500 - 12 mos. James A. Reggia University of Maryland at College Park Presidential Young Investigator Award: Abductive Inference Models in Artificial Intelligence - - - IST-8644983 $123,221 - 12 mos. Whitman A. Richards Massachusetts Institute of Technology Natural Computation: A Computational Approach to Visual Information Processing - - - IST-8604530 $49,953 - 12 mos. Fred S. Roberts Rutgers University Scales of Measurement and the Limitations they Place on Information Processing - - - IST-8603407 $27,000 - 12 mos. Robert D. Rodman North Carolina State University Dialog Processing for Voice Interactive Problem Solving - - - IST-8640925 $198,800 - 12 mos. Naomi Sager New York University Language As a Database Structure - - - IST-8640053 $59,178 - 12 mos. Sharon C. Salveter Boston University Transportable Natural Language Database Update - - - IST-8610293 $80,630 - 12 mos. Glenn R. Shafer University of Kansas Main Campus Belief Functions in Artificial Intelligence - - - IST-8603214 $85,583 - 12 mos. William Shaw University of North Carolina An Evaluation and Comparison of Term and Citation Indexing - - - DMS-8606178 $20,000 - 12 mos. Paul C. Shields University of Toledo Mathematical Sciences: Entropy in Ergodic Theory, Graph Theory and Statistics - - - IST-8607849 $101,839 - 12 mos. Edward Smith BBN Laboratories, Inc. A Computational Approach to Decision-Making - - - IST-8609599 $80,963 - 12 mos. Paul Smolensky University of Colorado at Boulder Inference in Massively Parallel Artificial Intelligence Systems - - - IST-8644907 $15,634 - 12 mos. Frederik Springsteel University of Missouri Formalization of Entity-Relationship Diagrams - - - IST-8640120 $66,004 - 12 mos. Robert E. Stepp University of Illinois at Urbana Discovering Underlying Concepts in Data Through Conceptual Clustering - - - IST-8516313 $60,907 - 12 mos. Richmond H. Thomason Mellon-Pitt-Carnegie Corp. Nonmonotonic Reasoning - - - IST-8516330 $63,854 - 12 mos. David S. Touretzky Carnegie-Mellon University Distributed Representations for Symbolic Data Structures - - - IST-8517289 $164,786 - 12 mos. Joseph F. Traub Columbia University The Information Level: Effective Computing with Partial, Contaminated, and Costly Information - - - IST-8544806 $121,222 - 12 mos. Jeffrey D. Ullman Stanford University Implementation of Logical Query Languages for Databases - - - IST-8511348 $30,383 - 12 mos. Kenneth Wexler University of California at Irvine Learnability and Parsability - - - IST-8514890 $80,000 - 12 mos. R. Wilensky and R. Alterman University of California at Berkeley Adaptive Planning - - - IST-8600788 $81,660 - 12 mos. Robert T. Winkler Duke University Combining Dependent Information: Models and Issues - - - IST-8644767 $38,474 - 12 mos. Ronald R. Yager Iona College Specificity Measures of Information in Possibility Distributions - - - IST-8644435 $50,878 - 12 mos. Po-Lung Yu University of Kansas Habitual Domain Analysis for Effective Information Interface and Decision Support - - - IST-8642900 $108,182 - 12 mos. Lotfi A. Zadeh University of California at Berkeley Management of Uncertainty in Expert Systems - - - IST-8605163 $19,776 - 12 mos. Maria Zemankova University of Tennessee at Knoxville Travel to the International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems - - - IST-8600616 $97,727 - 12 mos. Pranas Zunde Georgia Institute of Technology A Study of Word Association Aids in Information Retrieval ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Sun Nov 9 02:20:57 1986 Date: Sun, 9 Nov 86 02:20:51 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #253 Status: R AIList Digest Thursday, 6 Nov 1986 Volume 4 : Issue 253 Today's Topics: Funding - NSF Robotics and Machine Intelligence Awards ---------------------------------------------------------------------- Date: Fri 31 Oct 86 13:03:11-CST From: ICS.DEKEN@R20.UTEXAS.EDU Subject: Active Awards in the Robotics and Machine Intelligence Program (NSF) Fiscal Year 1986 Research Projects Funded by the Intelligent Systems Program (now Robotics and Machine Intelligence Program) A complete listing of these awards, including short descriptive abstracts of the research is available by writing to: Y.T. Chien, Director Robotics and Machine Intelligence Program National Science Foundation 1800 G Street NW Washington, DC 20550 ----- Computer Vision and Image Processing ----- SRI International; Alex P. Pentland; {Perceptual Organization and the Representation of Natural Scenes}; (DCR-8519283); $78060;12 months. Stanford University; Paul Switzer; {Statistical Theory and Methods for Processing Spatial Imagery (Mathematical Sciences and Computer Research)}; (DMS-8411300 A02); $15000; 12 months; (Joint support with the Statistics and Probability Program - Total Grant $48200). University of California - Berkeley; Alberto Grunbaum;{Reconstruction with Limited and Noisy Data (Mathematical Sciences and Computer Research)}; (DMS-8403232 A02); $8000; 12 months;(Joint support with the Applied Mathematics Program - Total Grant$66850). University of Miami; Tzay Y. Young; {Three-Dimensional Motion Analysis Using Shape Change Information (Computer Research)};(DCR-8509737 A01); $39871; 12 months. University of Illinois - Urbana; Thomas S. Huang; {Acquisition, Representation and Manipulation of Time-Varying Spatial Information (Computer Research)}; (DCR-8415325 A02); $14768; 6 months. University of Maryland - College Park; Azriel Rosenfeld;{Perceptual Organization in Computer Vision: Pyramid-based Approaches}; (DCR-8603723); $120376; 12 months. University of Maryland - College Park; Azriel Rosenfeld;{Workshop on Graph Grammars and their Application to Computer Science Leesburg Virginia October1986}; (DCR-8504840);$26235; 18 months. Massachusetts Institute of Technology; Whitman Richards;{Natural Computation: A Computational Approach to Visual Information Processing (Information Science and ComputerResearch)}; (IST-8312240 A02); $61500; 12 months; (Joint support with the Information Science Program - Total Grant $184721). Michigan State University; George C. Stockman and Anil K. Jain;{Feature Extraction and Evaluation in Recognition of 3D Objects};(DCR-8600371); $55522; 12 months. University of Michigan - Ann Arbor; Ramesh Jain; {Ego-Motion Complex Logarithmic Mapping}; (DCR-8517251); $48039; 12 months. University of Minnesota; William B. Thompson; {Determining Spatial Organization from Visual Motion (Computer Research)};(DCR-8500899 A01); $40750; 12 months. University of Rochester; Dana H. Ballard; {Parameter Networks and Spatial Cognition}; (DCR-8602958); $80464; 12 months. Carnegie-Mellon University; Steven A. Shafer and Takeo Kanade;{Optical Modeling in Image Understanding: Color Gloss and Shadows (Computer Research)}; (DCR-8419990 A01); $40475; 12 months. University of Wisconsin - Madison; Charles R. Dyer; {Parallel Vision Algorithms for Shared-Memory and Pipeline Multiprocessors};(DCR-8520870); $134857; 24 months. ----- Natural Language and Signal Understanding ----- SRI International; Douglas Appelt; {Natural Language Utterance Planning(Computer Research and Information Science)}; (DCR-8641243); $95210; 12 months; Indiana University - Bloomington; Robert F. Port Stan C. Kwasny and Daniel P. Maki; {Data-Driven Speech Recognition Using Prosody}; (DCR-8518725);$101055; 12 months. Massachusetts Institute of Technology; Robert C. Berwick; {PYI:(Computer Research)}; (DCR-8552543); $25000; 12 months. New York University; Ralph Grishman (in collaboration with Lynette Hirshman Buroughs Corporation); {Industry/University Cooperative Research: Acquisition and Use of Semantic Information for Natural Language Processing (ComputerResearch)}; (DCR-8501843 A01); $78400; 12 months; (Joint support with theIndustry/University Cooperative Research Program - Total Grant $93400). Duke University; Alan W. Biermann; {Dialog Processing for Voice Interactive Problem Solving}; (DCR-8603231); $27676; 12 months; (Joint support with the Special Projects Program the Information Science Program and the Information Technology Program - Total Grant $69676). North Carolina State University - Raleigh; Robert Rodman; {Dialogue Processing for Voice Interactive Problem Solving}; (DCR-8603407); $4000; 24months; (Joint support with the Special Projects Program the Information Science Program and the Information Technology Program - Total Grant $65276). Carnegie-Mellon University; Ronald A. Cole and Richard M. Stern; {Phonetic Classification =9Q%9U=UM Speech}; (DCR-8512695); $127523; 24 months. University of Pennsylvania; Aravind K. Joshi; {Research In Natural Language Processing (Computer Research)}; (DCR-8545755 A01); $160000; 12 months. Burroughs Corporation; Lynette Hirshman (in collaboration with Ralph Grishman New York University; {Industry/University Cooperative Research: Acquisition and Use of Semantic Information for Natural Language Processing(Computer Research)}; (DCR-8502205 A01); $58364; 12 months; (Joint support with the Industry/University Cooperative Research Program - Total Grant$73364). ----- Concept Learning and Inference ----- Yale University; Dana C. Angluin; {Algorithms for Inductive Inference (Computer Research)}; (DCR-8404226 A01); $88807; 12months. Northwestern University; Lawrence J. Henschen; {Logic and Databases}; (DCR-8608311); $44845; 12 months; (Joint support with the Information Science Program - Total Grant $89845). University of Chicago; James Royer; {Theory of Machine Learning}; (DCR-8602991); $17000; 24 months; (Joint support withTheoretical Computer Science Program - Total Grant $51045). University of Illinois - Urbana; R. S. Michalski; {Studies in Computer Inductive Learning and Plausible Inference}; (DCR-8645223A02); $135000; 12 months. University of Southwestern Louisiana; Rasiah Loganantharaj;{Theoretical and Implementational Aspects of Parallel Theorem Proving}; (DCR-8603039); $34994; 12 months. University of Maryland - College Park; Jack Minker; {Workshop on Foundations of Deductive Databases and Logic Programming College Park Maryland August 1986}; (DCR-8602676); $25390; 12 months. Rutgers University - Busch Campus; Tom M. Mitchell; {PYI:(Computer Research)}; (DCR-8351523 A03); $25000; 12 months. State University of New York - Albany; Neil V. Murray;{Automated Reasoning with Path Resolution and Semantic Graphs};(DCR-8600848); $34981; 12 months. Carnegie-Mellon University; Peter B. andrews; {Automated Theorem Proving in Type Theory (Computer Research)}; (DCR-8402532 A02);$87152; 12 months. Carnegie-Mellon University; Elaine Kant and Allen Newell;{Algorithm Design and Discovery (Computer Research)}; (DCR-8412139A01); $57328; 12 months. University of Texas - Austin; Michael P. Starbird and Woodrow W.Bledsoe; {Automatic Theorem Proving and Applications (ComputerResearch)}; (DCR-8313499 A02); $150930; 12 months. University of Wyoming; Michael J. Magee; {A Theorem Proving Based System for Recognizing Three-Dimensional Objects};(DCR-8602555); $37530; 12 months. ----- Knowledge Representation and Problem Solving ----- Stanford University; John McCarthy; {Artificial Intelligence (ComputerResearch)}; (DCR-8414393 A01); $134328; 12 months. Stanford University; Edward A. Feigenbaum and Charles Yanofsky;{MOLGEN-Applications of Artificial Intelligence to Molecular Biology Research in Theory Formation Testing and Modification (Computer Research)};(DCR-8310236 A02); $135000; 12 months. University of California - Berkeley; Lotfi A. Zadeh; {Fuzzy Logic as a Basis for Commonsense Reasoning and Inference in Expert Systems (Computer Research)};(DCR-8513139 A01); $100685; 12 months. University of California - Los Angeles; Judea Pearl; {Studies in Heuristics(Computer Research)}; (DCR-8501234 A01); $7895571; 12 months. University of California - Los Angeles; Judea Pearl and monthshe Ben-Bassat;{Toward a Computational Model of Evidential Reasoning (Computer Research)};(DCR-8313875 A02); $95291; 12 months. University of Southern California; Peter Waksman; {Grid Analysis - A Theory of Form Perception (Mathematical Sciences and Computer Research}; (DMS-8602025); $5000; 12 months (Joint support with the Applied Mathematics Program- Total Grant $15600). Yale University; Paul Hudak; {DAPS: Systems Support For AI (Computer Research)}; (DCR-8403304 A01); $62363; 12 months. University of Maryland - College Park; Dana S. Nau; {PYI: (ComputerResearch)}; (DCR-8351463 A02); $62500; 12 months. University of Maryland - College Park; Laveen N. Kanal; {Parallel Problem Solving and Applications in Artificial Intelligence}; (DCR-8504011 A01);$74922; 12 months. University of Maryland - College Park; Hanan Samet; {Hierarchical Data Structures}; (DCR-8605557); $45473; 12 months. University of Massachusetts - Amherst; Victor R. Lesser Krithivasan Ramamritham and Edward M. Riseman; {A Research Facility for Cooperative Distributed Computing (Computer Research)}; (DCR-8644692); $590977; 12 months;(Joint support with the Coordinated Experimental Research - Total Grant$984962). University of Michigan - Ann Arbor; Arthur W. Burks; {Languages and Architectures for Parallel Computing with Classifier Systems (ComputerResearch)}; (DCR-8305830 A03); $6381. University of Minnesota; James R. Slagle; {Expert Systems Questioning Procedures Based on Merit (Computer Research)}; (DCR-8512857 A01); $82235; 12months. University of New Hampshire; Eugene C. Freuder and Michael J. Quinn; {Copingwith Complexity in Constraint Satisfaction Problems}; (DCR-8601209); $32708;12 months; (Joint support with the Theoretical Computer Science Program - TotalGrant $42708). Rutgers University - Busch Campus; Saul Amarel and Charles Schmidt;{Exploration of Problem Reformulation and Strategy Acquisition}; (DCR-8318075A03); $73547; 12 months; (Joint support with the Information TechnologyProgram - Total Grant $102706). Rutgers University; Tomasz Imielinski; {Processing Incomplete Knowledge-A Database Approach (Computer Research)}; (DCR-8504140 A01); $57411; 12 months. New Mexico State University; Derek P. Partridge; {Workshop on the Foundations of Artificial Intelligence Las Cruces New Mexico February 1986};(DCR-8514964); $15000; 12 months. Cornell University; Robert L. Constable; {Experiments with a Program Refinement System (Computer Research)}; (DCR-8303327 A03); $60000; 12 months;(Joint support with the Software Engineering Program and the Software Systems Science Program - Total Grant $180247). Iona College; Ronald R. Yager; {Methods of Evidential Reasoning (ComputerResearch)}; (DCR-8513044); $38600; 12 months. University of Rochester; James F. Allen; {PYI: (Computer Research)};(DCR-8351665 A02); $25000; 12 months. University of Rochester; James F. Allen; {Temporal World Models for Problem Solving (Computer Research)}; (DCR-8502481 A01); $38375; 12 months. University of Texas - Austin; Benjamin J. Kuipers; {Deep and Shallow Models in the Knowledge Base}; (DCR-8602665); $45720; 12 months; (Joint support withthe Information Science Program - Total Grant $91440). ----- Automation and Robotics ----- Arizona State University; Kathleen M. Mutch; {Robotic Navigation Using Dynamic Imagery}; (DCR-8601798); $69955; 12 months. University of Illinois - Urbana; Thomas S. Huang; {Acquisition Representation and Manipulation of Time-Varying Spatial Information (ComputerResearch)}; (DCR-8640776 A01); $70752; 12 months. University of Massachusetts - Amherst; Edward M. Riseman and Arthur S.Gaylord; {A Group Research Facility for Artificial Intelligence Distributed Computing and Software Systems (Computer Research)}; (DCR-8318776 A03);$15000; 12 months; (Joint support with the Special Projects Program and the Software Engineering Program - Total Grant $80000). Cornell University-Endowed; John E. Hopcroft; {An International Workshop on Geometric Reasoning to be held June 30 - July 2 1986 at Keble College Oxford University U.K.}; (DCR-8605077); $22423; 12 months. Cornell University; John Hopcroft and Alan Demers; {A Program of Research in Robotics}; (DMC-8640765 A02); $91072; 12 months; Cornell University; John E. Hopcroft and Kuo K. Wang; {A Program of Researchin Representing Physical Objects}; (DCR-8644262 A01); $104238; 12 months. New York University; Ernest Davis; {Physical and Spatial Reasoning with Solid Objects}; (DCR-8603758 & A01); $82600; 24 months. New York University; David Lowe; {Model Based Recognition of Three-Dimensional Objects (Computer Research)}; (DCR-8502009 A01);$49700; 12 months. New York University; Colm O'Dunlaing and Chee-Keng Yap; {Motion Planning Problems in Robotics: Algorithmic Issues (Computer Research)}; (DCR-8401898A02); $97300; 12 months. Carnegie Mellon University; Takeo Kanade and Charles Thorpe; {Understanding 3-D Dynamic Natural Scenes with Range Data}; (DCR-8604199); $75817; 12 months. University of Pennsylvania; Ruzena K. Bajcsy; {Tactile Information Processing(Computer Research)}; (DCR-8545795); $45359; 12 months; University of Texas - Austin; J. K. Aggarwal; {Space Perception from Multiple Sensing}; (DCR-8517583); $125000; 24 months. University of Utah; Bir Bhanu and Thomas C. Henderson; {Computer Aided Geometric Design Based Computer Vision (Computer Research)}; (DCR-8644518);$74997; 12 months. ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Sun Nov 9 02:21:26 1986 Date: Sun, 9 Nov 86 02:21:18 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe.arpa Subject: AIList Digest V4 #254 Status: R AIList Digest Friday, 7 Nov 1986 Volume 4 : Issue 254 Today's Topics: Philosophy - Searle, Turing, Symbols, Categories ---------------------------------------------------------------------- Date: 28 Oct 86 19:54:22 GMT From: fluke!ssc-vax!bcsaic!michaelm@beaver.cs.washington.edu (michael maxwell) Subject: Re: Searle, Turing, Symbols, Categories In article <10@mind.UUCP> harnad@mind.UUCP (Stevan Harnad) writes: > >michaelm@bcsaic.UUCP (me) wrote: > >> As an interesting thought experiment, suppose a Turing test were done >> with a robot made to look like a human, and a human being who didn't >> speak English-- both over a CCTV, say, so you couldn't touch them to >> see which one was soft, etc. What would the robot have to do in order >> to pass itself off as human? > >...We certainly have no problem in principle with >foreign speakers (the remarkable linguist, polyglot and bible-translator >Kenneth Pike has a "magic show" in which, after less than an hour of "turing" >interactions with a speaker of any of the [shrinking] number of languages he >doesn't yet know, they are babbling mutually intelligibly before your very >eyes), although most of us may have some problems in practice with such a >feat, at least, without practice. Yes, you can do (I have done) such "magic shows" in which you begin to learn a language using just gestures + what you pick up of the language as you go along. It helps to have some training in linguistics, particularly field methods. The Summer Institute of Linguistics (of which Pike is President Emeritus) gives such classes. After one semester you too can give a magic show! I guess what I had in mind for the revised Turing test was not using language at all--maybe I should have eliminated the sound link (and writing). What in the way people behave (facial expressions, body language etc.) would cue us to the idea the one is a human and the other a robot? What if you showed pictures to the examinees--perhaps beautiful scenes, and revolting ones? This is more a test for emotions than for mind (Mr. Spock would probably fail). But I think that a lot of what we think of as human is tied up in this nonverbal/ emotional level. BTW, I doubt whether the number of languages Pike knows is shrinking because of these monolingual demonstrations (aka "magic shows") he's doing. After the tenth language, you tend to forget what the second or third language was-- much less what you learned! -- Mike Maxwell Boeing Advanced Technology Center ...uw-beaver!uw-june!bcsaic!michaelm ------------------------------ Date: 30 Oct 86 00:11:29 GMT From: mnetor!utzoo!utcsri!utegc!utai!me@seismo.css.gov Subject: Re: Searle, Turing, Symbols, Categories In article <1@mind.UUCP> harnad@mind.UUCP (Stevan Harnad) writes: >In reply to a prior iteration D. Simon writes: > >> I fail to see what [your "Total Turing Test"] has to do with >> the Turing test as originally conceived, which involved measuring >> up AI systems against observers' impressions, rather than against >> objective standards... Moreover, you haven't said anything concrete >> about what this test might look like. > >How about this for a first approximation: We already know, roughly >speaking, what human beings are able to "do" -- their total cognitive >performance capacity: They can recognize, manipulate, sort, identify and >describe the objects in their environment and they can respond and reply >appropriately to descriptions. Get a robot to do that. When you think >he can do everything you know people can do formally, see whether >people can tell him apart from people informally. > "respond and reply appropriately to descriptions". Very nice. Should be a piece of cake to formalize--especially once you've formalized recognition, manipulation, identification, and description (and, let's face it, any dumb old computer can sort). This is precisely what I was wondering when I asked you what this total Turing test looks like. Apparently, you haven't the foggiest idea, except that it would test roughly the same things that the old-fashioned, informal, does-it-look-smart-or-doesn't-it Turing test checks. In fact, none of the criteria you have described above seems defineable in any sense other than by reference to standard Turing test results ("gee, it sure classified THAT element the way I would've!"). And if you WERE to define the entire spectrum of human behaviour in an objective fashion ("rule 1: answering, 'splunge!' to any question is hereby defined as an 'appropriate reply'"), how would you determine whether the objective definition is useful? Why, build a robot embodying it, and see if people consider it intelligent, of course! The illusion of a "total" Turing test, distinct from the old-fashioned, subjective variety, thus vanishes in a puff of empiricism. And forget the well-that's-the-way-Science-does-it argument. It won't wash --see below. >> I believe that people in general dodge the "other minds" problem >> simply by accepting as a convention that human beings are by >> definition intelligent. > >That's an artful dodge indeed. And do you think animals also accept such >conventions about one another? Philosophers, at least, seem to >have noticed that there's a bit of a problem there. Looking human >certainly gives us the prima facie benefit of the doubt in many cases, >but so far nature has spared us having to contend with any really >artful imposters. Wait till the robots begin giving our lax informal >turing-testing a run for its money. > I haven't a clue whether animals think, or whether you think, for that matter. This is precisely my point. I don't believe we humans have EVER solved the "other minds" problem, or have EVER used the Turing test, even to try to resolve the question of whether there exist "other minds". The fact that you would like us to have done so, thus giving you a justification for the use of the (informal part of) the Turing test (and the subsequent implicit basing of the formal part on the informal part--see above), doesn't make it so. This is where your scientific-empirical model for developing the "total" Turing test out of the original falls down. Let's examine the development of a typical scientific concept: You have some rough, intuitive observations of phenomena (gravity, stars, skin). You take some objects whose properties you believe you understand (rocks, telescopes, microscopes), let them interact with your vaguely observed phenomenon, and draw more rigorous conclusions based on the recorded results of these experimental interactions. Now, let's examine the Turing test in that light: we take possibly-intelligent robot R, whose properties are fairly well understood, and sit it in front of person P, whose properties are something of a cipher to us. We then have them interact, and get a reading off person P (such as, "yup, shore is smart", or, "nope, dumb as a tree"). Now, what properties are being scientifically investigated here? They can't have anything to do with robot R--we assume that R's designer, Dr. Rstein, already has a fairly good idea what R is about. Rather, it appears as though you are discerning those attributes of people which relate to their judgment of intelligence in other objects. Of course, it might well turn out that something productive comes out of this, but it's also quite possible (and I conjecture that it's actually quite likely) that what you get out of this is some scientific law such as, "anything which is physically indistinguishable from a human being and can mutter something that sounds like person P's language is intelligent; anything else is generally dumb, but possibly intelligent, depending on the decoration of the room and the drug content of P's bloodstream at the time of the test". In short, my worries about the context-dependence and subjective quality of the results have not disappeared in a puff of empiricism; they loom as large as ever. > >> WHAT DOES THE "TOTAL TURING TEST" LOOK LIKE?... Please >> forgive my impertinent questions, but I haven't read your >> articles, and I'm not exactly clear about what this "total" >> Turing test entails. > >Try reading the articles. > Well, not only did I consider this pretty snide, but when I sent you mail privately, asking politely where I can find the articles in question, I didn't even get an answer, snide or otherwise. So starting with this posting, I refuse to apologize for being impertinent. Nyah, nyah, nyah. > > >Stevan Harnad >princeton!mind!harnad Daniel R. Simon "sorry, no more quotations" -D. Simon ------------------------------ Date: Thu, 30 Oct 86 16:09:20 EST From: "Col. G. L. Sicherman" Subject: Re: extended Turing test In article <8610271728.AA12616@ucbvax.Berkeley.EDU>, harnad@mind.UUCP writes: > > [I]t's misleading to propose that a veridical model of _our_ behavior > > ought to have our "performance capacities"...I do not (yet) quarrel > > with the principle that the model ought to have our abilities. But to > > speak of "performance capacities" is to subtly distort the fundamental > > problem. We are not performers! > > "Behavioral ability"/"performance capacity" -- such fuss over > black-box synonyms, instead of facing the substantive problem of > modeling the functional substrate that will generate them. You seem to be looking at the problem as a scientist. Let me give an example of what I mean: Suppose you have a robot slave. (That's the practical goal of A.I., isn't it?) It cooks for you, makes the beds, changes the oil in your car, puts the dog out, performs sexual favors, ... you name it. BUT-- it will not open the front door for you! Maddened with frustration, you order an electric-eye door opener, 1950s design. It works flawlessly. Now you have everything you want. Does the combination of robot + door-opener pass the Total Turing Test? Is the combination a valid subject for the Test? ------------------------------ Date: Thu, 30 Oct 86 15:56:27 EST From: "Col. G. L. Sicherman" Subject: Re: how we decide whether it has a mind In article <8610271726.AA12550@ucbvax.Berkeley.EDU>, harnad@mind.UUCP writes: > > One (rationally) believes other people are conscious BOTH because > > of their performance and because their internal stuff is a lot like > > one's own. > > ... I am not denying that > there exist some objective data that correlate with having a mind > (consciousness) over and above performance data. In particular, > there's (1) the way we look and (2) the fact that we have brains. What > I am denying is that this is relevant to our intuitions about who has a > mind and why. I claim that our intuitive sense of who has a mind is > COMPLETELY based on performance, and our reason can do no better. ... There's a complication here: Our intutions about things in our environment change with the environment. The first time you use a telephone, you hear an electronic reproduction of somebody's voice; you KNOW that you're talking to a machine, not to the other person. Soon this knowledge evaporates, and you come to think, "I talked with Smith today on the phone." You may even have seen his face before you! It's the same with thinking. When only living things could perceive and adapt accordingly, people did not think of artifacts as having minds. This wasn't stubborn of them, just honest intuition. When ELIZA came along, it became useful for her users to think of her as having a mind. Just like thinking you talked with Smith ... I'd like to see less treatment of "X has a mind" as a formal proposition, and more discussion of how we use our intuition about it. After all, is having a mind the most important thing about somebody to you? Is it even important at all? ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Sun Nov 9 02:21:40 1986 Date: Sun, 9 Nov 86 02:21:31 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe Subject: AIList Digest V4 #255 Status: R AIList Digest Friday, 7 Nov 1986 Volume 4 : Issue 255 Today's Topics: Philosophy - Searle, Turing, Symbols, Categories ---------------------------------------------------------------------- Date: 29 Oct 86 18:31:18 GMT From: ubc-vision!ubc-cs!andrews@BEAVER.CS.WASHINGTON.EDU Subject: Turing Test ad infinitum This endless discussion about the Turing Test makes the "eliminative materialist" viewpoint very appealing: by the time we have achieved something that most people today would call intelligent, we will have done it through disposing of concepts such as "intelligence", "consciousness", etc. Perhaps the reason we're having so much trouble defining a workable Turing Test is that we're essentially trying to fit a square peg into a round hole, belabouring some point which has less relevance than we realize. I wonder what old Alan himself would say about the whole mess. --Jamie. ...!seismo!ubc-vision!ubc-cs!andrews "At the sound of the falling tree... it's 9:30" ------------------------------ Date: 1 Nov 86 20:34:02 GMT From: allegra!princeton!mind!harnad@ucbvax.Berkeley.EDU (Stevan Harnad) Subject: Re: Searle, Turing, Symbols, Categories In his second net.ai comment on the abstracts of the two articles under discussion, me@utai.UUCP (Daniel Simon) wrote: >> WHAT DOES THE "TOTAL TURING TEST" LOOK LIKE?... Please >> forgive my impertinent questions, but I haven't read your >> articles, and I'm not exactly clear about what this "total" >> Turing test entails. I replied (after longish attempts to explain in two separate iterations): >"Try reading the articles." Daniel Simon rejoined: > Well, not only did I consider this pretty snide, but when I sent you > mail privately, asking politely where I can find the articles in > question, I didn't even get an answer, snide or otherwise. So starting > with this posting, I refuse to apologize for being impertinent. > Nyah, nyah, nyah. The same day, the following email came from Daniel Simon: > Subject: Hoo, boy, did I put my foot in it: > Ooops....Thank you very much for sending me the articles, and I'm sorry > I called you snide in my last posting. If you see a bright scarlet glow > in the distance, looking west from Princeton, it's my face. Serves me > right for being impertinent in the first place... As soon as I finish > reading the papers, I'll respond in full--assuming you still care what > I have to say... Thanks again. Yours shamefacedly, Daniel R. Simon. This is a very new form of communication for all of us. We're just going to have to work out a new code of Nettiquette. With time, it'll come. I continue to care what anyone says with courtesy and restraint, and intend to respond to everything of which I succeed in making sense. Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ Date: 1 Nov 86 18:21:12 GMT From: allegra!princeton!mind!harnad@ucbvax.Berkeley.EDU (Stevan Harnad) Subject: Re: Searle, Turing, Symbols, Categories Jay Freeman (freeman@spar.UUCP) had, I thought, joined the ongoing discussion about the robotic version of the Total Turing Test to address the questions that were raised in the papers under discussion, namely: (1) Do we have any basis for contending with the "other minds problem" -- whether in other people, animals or machines -- other than turing-indistinguishable performance capacity? (2) Is the teletype version of the turing test -- which allows only linguistic (i.e., symbolic) interactions -- a strong enough test? (3) Could even the linguistic version alone be successfully passed by any device whose symbolic functions were not "grounded" in nonsymbolic (i.e., robotic) function? (4) Are transduction, analog representations, A/D conversion, and effectors really trivial in this context, or is there a nontrivial hybrid function, grounding symbolic representation in nonsymbolic representation, that no one has yet worked out? When Freeman made his original sugestion that the symbolic processor could have access to the robotic transducer's bit-map, I thought he was making the sophisticated (but familiar) point that once the transducer representation is digitized, it's symbolic all the way. (This is a variant of the "transduction-is-trivial" argument.) My prior reply to Freeman (about simulated models of the world, modularity, etc.) was addressed to this construal of his point. But now I see that he was not making this point at all, for he replies: > ... let's equip the robot with an active RF emitter so > it can jam the camera's electronics and impose whatever bit map it > wishes... design a robot in the shape of a back projector, and let it > create internally whatever representation of a human being it wishes > the camera to see, and project it on its screen for the camera to > pick up. Such a robot might do a tolerable job of interacting with > other parts of the "objective" world, using robot arms and whatnot > of more conventional design, so long as it kept them out of the > way of the camera... let's create a vaguely anthropomorphic robot and > equip its external surfaces with a complete covering of smaller video > displays, so that it can achieve the minor details of human appearance > by projection rather than by mechanical motion. Well, maybe our model > shop is good enough to do most of the details of the robot convincingly, > so we'll only have to project subtle details of facial expression. > Maybe just the eyes. > ... if you are going to admit the presence of electronic or mechanical > devices between the subject under test and the human to be fooled, > you must accept the possibility that the test subject will be smart > enough to detect their presence and exploit their weaknesses... > consider a robot that looks no more anthropomorphic than your vacuum > cleaner, but that is possessed of moderate manipulative abilities and > a good visual perceptive apparatus. > Before the test commences, the robot sneakily rolls up to the > camera and removes the cover. It locates the connections for the > external video output, and splices in a substitute connection to > an external video source which it generates. Then it replaces the > camera cover, so that everything looks normal. And at test time, > the robot provides whatever image it wants the testers to see. > A dumb robot might have no choice but to look like a human being > in order to pass the test. Why should a smart one be so constrained? >From this reply I infer that Freeman is largely concerned with the question of appearance: Can a robot that doesn't really look like a person SIMULATE looking like a person by essentially symbolic means, plus add-on modular peripherals? In the papers under discussion (and in some other iterations of this discussion on the net) I explicitly rejected appearance as a criterion. (The reasons are given elsewhere.) What is important in the robotic version is that it should be a human DO-alike, not a human LOOK-alike. I am claiming that the (Total) object-manipulative (etc.) performance of humans cannot be generated by a basically symbolic module that is merely connected with peripheral modules. I am hypothesizing (a) that symbolic representations must be NONMODULARLY (i.e., not independently) grounded in nonsymbolic representations, (b) that the Total Turing Test requires the candidate to display all of our robotic capacities as well as our linguistic ones, and (c) that even the linguistic ones could not be accomplished unless grounded in the robotic ones. In none of this do the particulars of what the robot (or its grey matter!) LOOK like matter. Two last observations. First, what the "proximal stimulus" -- i.e., the physical energy pattern on the transducer surface -- PRESERVES whereas the next (A/D) step -- the digital representation -- LOSES, is everything about the full PHYSICAL configuration of the energy pattern that cannot be recovered by inversion (D/A). (That's what the ongoing concurrent discussion about the A/D distinction is in part concerned with.) Second, I think there is a tendency to overcomplicate the issues involved in the turing test by adding various arbitrary elaborations to it. The basic questions are fairly simply stated (though not so simple to answer). Focusing instead on ornamented variants often seems to lead to begging the question or changing the subject. Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ Date: 1 Nov 86 20:02:08 GMT From: rutgers!princeton!mind!harnad@lll-crg.arpa (Stevan Harnad) Subject: Re: Searle, Turing, Symbols, Categories michaelm@bcsaic.UUCP (michael maxwell) writes: > I guess what I had in mind for the revised Turing test was not using > language at all--maybe I should have eliminated the sound link (and > writing). What in the way people behave (facial expressions, body > language etc.) would cue us to the idea the one is a human and the other > a robot? What if you showed pictures to the examinees--perhaps > beautiful scenes, and revolting ones? This is more a test for emotions > than for mind (Mr. Spock would probably fail). But I think that a lot of > what we think of as human is tied up in this nonverbal/ emotional level. The modularity issue looms large again. I don't believe there's an independent module for affective expression in human beings. It's all -- to use a trendy though inadequate expression -- "cognitively penetrable." There's also the issue of the TOTALITY of the Total Turing Test, which was intended to remedy the underdetermination of toy models/modules: It's not enough just to get a model to mimic our facial expressions. That could all be LITERALLY done with mirrors (and, say, some delayed feedback and some scrambling and recombining), and I'm sure it could fool people, at least for a while. I simply conjecture that this could not be done for the TOTALITY of our performance capacity using only more of the same kinds of tricks (analog OR symbolic). The capacity to manipulate objects in the world in all the ways we can and do do it (which happens to include naming and describing them, i.e., linguistic acts) is a lot taller order than mimicking exclusively our nonverbal expressive behavior. There may be (in an unfortunate mixed metaphor) many more ways to skin (toy) parts of the theoretical cat than all of it. Three final points: (1) Your proposal seems to equivocate between the (more important) formal functional component of the Total Turing Test (i.e., how do we get a model to exhibit all of our performance capacities, be they verbal or nonverbal?) and (2) the informal, intuitive component (i.e., will it be indistinguishable in all relevant respects from a person, TO a person?). The motto would be: If you use something short of the Total Turing Test, you may be able to fool some people some of the time, but not all of the time. (2) There's nothing wrong in principle with a nonverbal, even a nonhuman turing test; I think (higher) animals pass this easily all the time, with virtually the same validity as humans, as far as I'm concerned. But this version can't rely exclusively on affective expression modules either. (3) Finally, as I've argued earlier, all attempts to "capture" qualitative experience -- not just emotion, but any conscious experience, such as what it's LIKE to see red or to believe X -- amounts to an unprofitable red herring in this enterprise. The whole point of the Total Turing Test is that performance-indistinguishability IS your only basis for infer that anyone but you has a mind (i.e., has emotions, etc.). In the paper I dubbed this "methodological epiphenomenalism as aresearch strategy in cognitive science." By the way, you prejudged the question the way you put it. A perfectly noncommittal but monistic way of putting it would be: "What in the way ROBOTS behave would cue us to the idea that one robot had a mind and another did not?" This leaves it appropriately open for continuing research just exactly which causal physical devices (= "robots"), whether natural or artificial, do or do not have minds. Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Sun Nov 9 02:22:01 1986 Date: Sun, 9 Nov 86 02:21:50 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe Subject: AIList Digest V4 #256 Status: R AIList Digest Friday, 7 Nov 1986 Volume 4 : Issue 256 Today's Topics: Philosophy - Searle, Turing, Symbols, Categories ---------------------------------------------------------------------- Date: 2 Nov 86 23:22:23 GMT From: allegra!princeton!mind!harnad@ucbvax.Berkeley.EDU (Stevan Harnad) Subject: Re: Searle, Turing, Symbols, Categories The following is a response on net.ai to a comment on mod.ai. Because of problems with posting to mod.ai, I am temporarily replying to net.ai. On mod.ai cugini@NBS-VMS.ARPA ("CUGINI, JOHN") writes: > You seem to want to pretend that we know absolutely nothing about the > basis of thought in humans, and to "suppress" all evidence based on > such knowledge. But that's just wrong. Brains *are* evidence for mind, > in light of our present knowledge. What I said was that we knew absolutely nothing about the FUNCTIONAL basis of thought in humans, i.e., about how brains or relevantly similar devices WORK. Hence we wouldn't have the vaguest idea if a given lump of grey matter was in fact the right stuff, or just a gelatenous look-alike -- except by examining its performance (i.e., turing) capacity. [The same is true, by the way, mutatis mutandis, for a better structural look-alike -- with cells, synapses, etc. We have no functional idea of what differentiates a mind-supporting look-alike from a comatose one, or one from a nonviable fetus. Without the performance criterion the brain cue could lead us astray as often as not regarding whether there was indeed a mind there. And that's not to mention that we knew perfectly well (perhaps better, even) how to judge whether somebody had a mind before 'ere we ope'd a skull nor knew what we had chanced upon there. If you want a trivial concession though, I'll make one: If you saw an inert body totally incapable of behavior, then or in the future, and you entertained some prior subjective probability that it had a mind, say, p, then, if you opened its skull and found something anatomically and physiologically brain-like in there, then the probability p that it had, or had had, a mind would correspondingly rise. Ditto for an inert alien species. And I agree that that would be rational. However, I don't think that any of that has much to do with the problem of modeling the mind, or with the relative strengths or weaknesses of the Total Turing Test. > People in, say, 1500 AD were perfectly rational in predicting > tides based on the position of the moon (and vice-versa) > even though they hadn't a clue as to the mechanism of interaction. > If you keep asking "why" long enough, *all* science is grounded on > such brute-fact correlation (why do like charges repel, etc.) - as > Hume pointed out a while back. Yes, but people then and even earlier were just as good at "predicting" the presence of mind WITHOUT any reference to the brain. And in ambiguous cases, behavior was and is the only rational arbiter. Consider, for example, which way you'd go if (1) an alien body persisted in behaving like a clock-like automaton in every respect -- no affect, no social interaction, just rote repetition -- but it DID have something that was indistinguishable (on the minute and superficial information we have) from a biological-like nervous system), versus (2) if a life-long close friend of yours had to undergo his first operation, and when they opened him up, he turned out to be all transistors on the inside. I don't set much store by this hypothetical sci-fi stuff, especially because it's not clear whether the "possibilities" we are contemplating are indeed possible. But the exercise does remind us that, after all, performance capacity is our primary criterion, both logically and intuitively, and its black-box correlates have whatever predictive power they may have only as a secondary, derivative matter. They depend for their validation on the behavioral criterion, and in cases of conflict, behavior continues to be the final arbiter. I agree that scientific inference is grounded in observed correlations. But the primary correlation in this special case is, I am arguing, between mental states and performance. That's what both our inferences and our intuitions are grounded in. The brain correlate is an additional cue, but only inasmuch as it agrees with performance. As to CAUSATION -- well, I'm sceptical that anyone will ever provide a completely satisfying account of the objective causes of subjective effects. Remember that, except for the special case of the mind, all other scientific inferences have only had to account for objective/objective correlations (and [or, more aptly, via) their subjective/subjective experiential counterparts). The case under discussion is the first (and I think only) case of objective/subjective correlation and causation. Hence all prior bets, generalizations or analogies are off or moot. > other brains... are, by definition, relevantly brain-like I'd be interested in knowing what current definition will distinguish a mind-supporting brain from a non-mind-supporting brain, or even a pseudobrain. (That IS the issue, after all, in claiming that the brain in an INDEPENDENT predictor of mindedness.) > Let me re-cast Harnad's argument (perhaps in a form unacceptable to > him): We can never know any mind directly, other than our own, if we > take the concept of mind to be something like "conscious intelligence" - > ie the intuitive (and correct, I believe) concept, rather than > some operational definition, which has been deliberately formulated > to circumvent the epistemological problems. (Harnad, to his credit, > does not stoop to such positivist ploys.) But the only external > evidence we are ever likely to get for "conscious intelligence" > is some kind of performance. Moreover, the physical basis for > such performance will be known only contingently, ie we do not > know, a priori, that it is brains, rather than automatic dishwashers, > which generate mind, but rather only as an a posteriori correlation. > Therefore, in the search for mind, we should rely on the primary > criterion (performance), rather than on such derivative criteria > as brains. I pretty much agree with the above account except for the > last sentence which prohibits us from making use of derivative > criteria. Why should we limit ourselves so? Since when is that part > of rationality? I accept the form in which you've recast my argument. The reasons that brainedness is not a good criterion are the following (I suppose I should stop saying it is not a "rational" criterion having made the minor concession I did above): Let's call being able to pass the Total Turing Test the "T" correlate of having a mind, and let's call having a brain the "B" correlate. (1) The validity of B depends completely on T. We have intuitions about the way we and others behave, and what it feels like; we have none about having brains. (2) In case of conflict between T and B, our intuitions (rationally, I suggest) go with T rather than B. (3) The subjective/objective issue (i.e., the mind/body problem) mentioned above puts these "correlations" in a rather different category from other empirical correlations, which are uniformly objective/objective. (4) Looked at sufficiently minutely and functionally, we don't know what the functionally relevant as opposed to the superficial properties of a brain are, insofar as mind-supportingness is concerned; in other words, we don't even know what's a B and what's just naively indistinguishable from a B (this is like a caricature of the turing test). Only T will allow us to pick them out. I think those are good enough reasons for saying that B is not a good independent criterion. That having be said, let me concede that for a radical sceptic, neither is T, for pretty pretty much the same reasons! This is why I am a methodological epiphenomenalist. > No, the fact is we do have more reason to suppose mind of other > humans than of robots, in virtue of an admittedly derivative (but > massively confirmed) criterion. And we are, in this regard, in an > epistemological position *superior* to those who don't/didn't know > about such things as the role of the brain, ie we have *more* reason > to believe in the mindedness of others than they do. That's why > primitive tribes (I guess) make the *mistake* of attributing > mind to trees, weather, etc. Since raw performance is all they > have to go on, seemingly meaningful activity on the part of any > old thing can be taken as evidence of consciousness. But we > sophisticates have indeed learned a thing or two, in particular, that > brains support consciousness, and therefore we (rationally) give the > benefit of the doubt to any brained entity, and the anti-benefit to > un-brained entities. Again, not to say that we might not learn about > other bases for mind - but that hardly disparages brainedness as a > rational criterion for mindedness. A trivially superior position, as I've suggested. Besides, the primitive's mistake (like the toy AI-modelers') is in settling for anything less than the Total Turing Test; the mistake is decidedly NOT the failure to hold out for the possession of a brain. I agree that it's rational to take brainedness as an additional corroborative cue, if you ever need one, but since it's completely useless when it fails to corroborate or conflicts with the Total Turing criterion, of what independent use is it? Perhaps I should repeat that I take the context for this discussion to be science rather than science fiction, exobiology or futurology. The problem we are presumably concerned with is that of providing an explanatory model of the mind along the lines of, say, physics's explanatory model of the universe. Where we will need "cues" and "correlates" is in determining whether the devices we build have succeeded in capturing the relevant functional properties of minds. Here the (ill-understood) properties of brains will, I suggest, be useless "correlates." (In fact, I conjecture that theoretical neuroscience will be led by, rather than itself leading, theoretical "mind-science" [= cognitive science?].) In sci-fi contexts, where we are guessing about aliens' minds or those of comatose creatures, having a blob of grey matter in the right place may indeed be predictive, but in the cog-sci lab it is not. > there's really not much difference between relying on one contingent > correlate (performance) rather than another (brains) as evidence for > the presence of mind. To a radical sceptic, as I've agreed above. But there is to a working cognitive scientist (whose best methodological stance, I suggest, is epiphenomenalism). > I know consciousness (my own, at least) exists, not as > some derived theoretical construct which explains low-level data > (like magnetism explains pointer readings), but as the absolutely > lowest rock-bottom datum there is. Consciousness is the data, > not the theory - it is the explicandum, not the explicans (hope > I got that right). It's true that I can't directly observe the > consciousness of others, but so what? That's an epistemological > inconvenience, but it doesn't make consciousness a red herring. I agree with most of this, and it's why I'm not, for example, an "eliminative materialist." But agreeing that consciousness is data rather than theory does not entail that it's the USUAL kind of data of empirical science. I KNOW I have a mind. Every other instance is radically different from this unique one: I can only guess, infer. Do you know of any similar case in normal scientific inference? This is not just an "epistemological inconvenience," it's a whole 'nother ball game. If we stick to the standard rules of objective science (which I recommend), then turing-indistinguishable performance modeling is indeed the best we can aspire to. And that does make consciousness a red herring. > ...being-composed-of-protein might not be as practically incidental > as many assume. Frinstance, at some level of difficulty, one can > get energy from sunlight "as plants do." But the issues are: > do we get energy from sunlight in the same way? How similar do > we demand that the processes are?...if we're interested in simulation at > a lower level of abstraction, eg, photosynthesis, then, maybe, a > non-biological approach will be impractical. The point is we know we > can simulate human chess-playing abilities with non-biological > technology. Should we just therefore declare the battle for mind won, > and go home? Or ask the harder question: what would it take to get a > machine to play a game of chess like a person does, ie, consciously. This sort of objection to a toy problem like chess (an objection I take to be valid) cannot be successfully redirected at the Total Turing Test, and that was one of the central points of the paper under discussion. Nor are the biological minutiae of modeling plant photosynthesis analogous to the biological minutiae of modeling the mind: The OBJECTIVE data in the mind case are what you can observe the organism to DO. Photosynthesis is something a plant does. In both cases one might reasonably demand that a veridical model should mimic the data as closely as possible. Hence the TOTAL Turing Test. But now what happens when you start bringing in physiological data, in the mind case, to be included with the performance data? There's no duality in the case of photosynthesis, nor is there any dichotomy of levels. Aspiring to model TOTAL photosynthesis is aspiring to get every chemical and temporal detail right. But what about the mind case? On the one hand, we both agree with the radical sceptic that NEITHER mimicking the behavior NOR mimicking the brain can furnish "direct" evidence that you've captured mind. So whereas getting every (observable) photosynthetic detail right "guarantees" that you've captured photosynthesis, there's no such guarantee with consciousness. So there's half of the disanalogy. Now consider again the hypothetical possibilities we were considering earlier: What if brain data and behavioral data compete? Which way should a nonsceptic vote? I'd go with behavior. Besides, it's an empirical question, as I said in the papers under discussion, whether or not brain constraints turn out to be relevant on the way to Total Turing Utopia. Way down the road, after all, the difference between mind-performance and brain-performance may well become blurred. Or it may not. I think the Total Turing Test is the right provisional methodology for getting you there, or at least getting you close enough. The rest may very well amount to only the "fine tuning." > BTW, I quite agree with your more general thesis on the likely > inadequacy of symbols (alone) to capture mind. I'm glad of that. But I have to point out that a lot of what you appear to disagree about went into the reasons supporting that very thesis, and vice versa. ----- May I append here a reply to andrews@ubc-cs.UUCP (Jamie Andrews) who wrote: > This endless discussion about the Turing Test makes the > "eliminative materialist" viewpoint very appealing: by the > time we have achieved something that most people today would > call intelligent, we will have done it through disposing of > concepts such as "intelligence", "consciousness", etc. > Perhaps the reason we're having so much trouble defining > a workable Turing Test is that we're essentially trying to > fit a square peg into a round hole, belabouring some point > which has less relevance than we realize. I wonder what old > Alan himself would say about the whole mess. On the contrary, rather than disposing of them, we will finally have some empirical and theoretical idea of what their functional basis might be, rather than simply knowing what it's like to have them. And if we don't first sort out our methodological constraints, we're not headed anywhere but in hermeneutic circles. Stevan Harnad {allegra, bellcore, seismo, rutgers, packard} !princeton!mind!harnad harnad%mind@princeton.csnet (609)-921-7771 ------------------------------ Date: 3 Nov 86 17:41:54 GMT From: mcvax!ukc!warwick!rlvd!kgd@seismo.css.gov (Keith Dancey) Subject: Re: Searle, Turing, Symbols, Categories In article <5@mind.UUCP> harnad@mind.UUCP (Stevan Harnad) writes: > > >What do you think "having intelligence" is? Turing's criterion >effectively made it: having performance capacity that is indistinguishable >from human performance capacity. And that's all "having a mind" >amounts to (by this objective criterion). ... At the risk of sidetracking this discussion, I don't think it wise to try and equate 'mind' and 'intelligence'. A 'mind' is an absolute thing, but 'intelligence' is relative. For instance, most people would, I believe, accept that a monkey has a 'mind'. However, they would not necessarily so easily accept that a monkey has 'performance capacity that is indistinguishable from human performance capacity'. On the other hand, many people would accept that certain robotic processes had 'intelligence', but would be very reluctant to attribute them with 'minds'. I think there is something organic about 'minds', but 'intelligence' can be codified, within limits, of course. I apologise if this appears as a red-herring in the argument. -- Keith Dancey, UUCP: ..!mcvax!ukc!rlvd!kgd Rutherford Appleton Laboratory, Chilton, Didcot, Oxon OX11 0QX JANET: K.DANCEY@uk.ac.rl Tel: (0235) 21900 ext 5716 ------------------------------ End of AIList Digest ******************** From in%@vtcs1 Sun Nov 9 02:22:10 1986 Date: Sun, 9 Nov 86 02:22:02 est From: in%@vtcs1 (LAWS@SRI-STRIPE.ARPA) To: ailist@sri-stripe Subject: AIList Digest V4 #257 Status: R AIList Digest Saturday, 8 Nov 1986 Volume 4 : Issue 257 Today's Topics: Administrivia - Net.ai is Being Renamed Comp.aim Queries - TCPIP Between Symbolics, Xerox 1100s and VAXs & Connectionism & Neural Modeling, Learning - Boltzmann Machines and Simulated Annealing, Expert Systems - Performance Analysis/Tuning Query, Education - Cognitive Science Grad Schools, Logic - Nonmonotonic Reasoning, Literature - Sentient-Computer Novels & The AI Gang ---------------------------------------------------------------------- Date: 7 Nov 86 20:01:25 GMT From: cbosgd!mark@ucbvax.Berkeley.EDU (Mark Horton) Subject: net.ai is being renamed comp.ai This newsgroup is being renamed from net.ai to comp.ai. This renaming will gradually take place over the next few weeks. More and more messages posted to this newsgroup will be aliased into the new newsgroup as they pass through the net, and people will begin to post to the new group. After a few weeks, the old name will be removed. This note is to inform you of the renaming so you can begin to read the new group as well as the old group. Mark Horton Director, the UUCP Project ------------------------------ Date: 4 Nov 86 19:29:27 GMT From: ihnp4!ihwpt!clarisse@ucbvax.Berkeley.EDU (olivier clarisse) Subject: Do YOU TCPIP between: Symbolics, Xerox 1100's and VAX's? Does anyone of you AI WORKSTATION USERS work on a local network (ethernet) running TCPIP and have used SYMBOLICS 3600s and (or) VAXes running UNIX (system 5 for example) as hosts for FTP to XEROX 1186 (or 110X)? IF YOU HAVE experienced such things, please let me now how it goes: great or terrible? Is the communication smooth (or like a dotted line?) Do you use the SYMBOLICS (VAX) as a file server? Have you purchased a software to be able to use the 110X as a host too? Which one? (The 110X as a host is not supported on TCPIP by XEROX, I just heard, while FTP is supposed if someone else is the host.) Please let me know about your exciting experiences with TCPIP/FTP and be as specific as possible with respect to the system/software used. THANKS AN 1186 TIMES! Olivier Clarisse clarisse@ihesa@ihnp4.uucp (312) 979-3558 ------------------------------ Date: 7 Nov 86 22:06:21 GMT From: mcvax!cernvax!ethz!wyle@seismo.css.gov (Mitchell Wyle) Subject: Connectionism, neural networks: new mail list or group? Is anyone interested in a net.connectionism group? What about a mailing list? If anyone is interested in contributing to or receiving a tentative bibliography of connectionism/neural nets, let me know. ~~~~~~~~~~~~~~~~~~~~~~ Mitch Wyle ...!decvax!seismo!mcvax!cernvax!ethz!Wyle Institut fuer Informatik Wyle%ifi.ethz.cernvax. ETH / SOT 8092 Zuerich, Switzerland "Ignore alien orders." ------------------------------ Date: 4 Nov 86 11:25:18 GMT From: mcvax!ukc!dcl-cs!strath-cs!jrm@seismo.css.gov (Jon R Malone) Subject: Request for information (Brain/Parallel fibers) <<<>>> Nice guy, into brains would like to meet similiarly minded people. Seriously : considering some simulation of neural circuits. Would like pointers to any REAL work that is going on (PS I have read the literature). Keen to run into somebody that is interested in simulation at a low-level. Specifically: * mossy fibers/basket cells/purkyne cells * need to find out parallel fiber details: * length of * source of/destination of Any pointers or info would be appreciated. ------------------------------ Date: 4 Nov 86 18:40:15 GMT From: mcvax!ukc!stc!datlog!torch!paul@seismo.css.gov (paul) Subject: Re: THINKING COMPUTERS ARE A REALITY (?) People who read the original posting in net.general (and the posting about neural networks in this newsgroup) may be interested in the following papers: Boltzmann Machines: Constraint Satisfaction Networks that Learn. by Geoffrey E. Hinton, Terrence J. Sejnowski and David H. Ackley Technical Report CMU-CS-84-119 (Carnegie-Mellon University May 1984) Optimisation by Simulated Annealing by S. Krikpatrick, C.D.Gelatt Jr., M.P.Vecchi Science Vol. 220 No. 4598 (13th May 1983). ...in addition to those recommended by Jonathan Marshall. Personally I regard this type of machine learning as something of a holy grail. In my opinion (and I stress that it IS my own opinion) this is THE way to get machines that are both massively parallel and capable of complex tasks without having a programmer who understands the in's and out's of the task to be accomplished and who is prepared to spend the time to hand code (or design) the machine necessary to do it. The only reservation I have is whether or not the basic theory behind Boltzmann machines is good enough. Paul. ------------------------------ Date: Fri 7 Nov 86 12:30:49-MST From: Sue Tabron Subject: expert systems for performance analysis/tuning? I am interested in finding public domain or commercial expert systems that can be used to analyze and tune the performance of computer systems. I would like to hear from anyone with experience in this area or who is developing these applications.. Mott Given (614)238-9431 ------------------------------ Date: 6 Nov 86 17:51:11 GMT From: tektronix!orca!tekecs!mikes@ucbvax.Berkeley.EDU (Michael Sellers) Subject: choosing grad schools [Note: This is a new subject. The words "Turing", "Searle", "ducks", and "categories" do not appear in this posting...okay, so I lied :-)] There was a little discussion some time ago regarding grad programs in cognitive science. Well, its that time of year when I begin to dream about selling the house and going for the old P, h, & D. So: For those of you who are in doctrate programs (or master's programs, too) in cognitive science, how did you choose the program you're in? What do you like/dislike about it? What are your employment prospects when you're done? What sorts of things drove your decision of what school to go to? What is your personal situation (single/married x number of kids, x years work experience, etc)? What I'm hoping to get is an idea of what the various programs are like from the inside; I can get all the propoganda I can stomach from various admissions offices. Thanks for your help. Post or e-mail as you want; if there is a lot of mail I'll summarize and post it. -- Mike Sellers UUCP: {...your spinal column here...}!tektronix!tekecs!mikes "In a quiet moment, you can just hear them brain cells a-dyin'" ------------------------------ Date: 7 Nov 86 15:34:47 GMT From: sdics!norman@sdcsvax.ucsd.edu (Donald A. Norman) Subject: Re: choosing grad schools (Weird that so many Cognitive Science issues end up in the Cognitive Engineering and AI newsgroups. Cog-Eng was originally human-computer interaction (the engineering, applied side of studies of cognition). As for AI, well, the part that deals with the understanding and simulation of thought is a subset of Cognitive Science, so it belongs.) Grad schools in Cognitive Science. I would ike to hear a summary (from knowledgable folks) of what exists. Here is what I know. There are NO departments of Cognitive Science. I know of only three places that offer degrees that have the phrase "Cognitive Science" in them (and 3 more that might, but I am not sure). The three I know of are Brwon, MIT, and UC San Diego (UCSD). The three I am not sure about are Rochester, SUNY Buffalo, and UC, Berkeley (UCB). Brown has a department of Linguistics and Cognitive Science. MIT has a department of Brain and Cognitive Science. UCSD has a "program in Cognitive Science" that offers a degree that reads "PhD in X and Cogntive Science", where X is one of the particpating departments (Anthropology, Computer Science, Linguistics, Music, Neuroscience, Philosophy, Psychology, Sociology) Rochester, SUNY Buffalo, and UCB have programs that might also offer some sort of degree, but I am not certain. Many other places have research programs in Cognitive Science, but as far as I know, no degree program. The UCSD program, for example, does not admit students directly into the program (we can't: we are not a department). Students apply to and are admitted into one of the coperating departments. At the end of the first year of stuidy, they apply to and enter the joint program with Cog Sci. At the end, the degree reads "PhD in X and Cognitive Science." There is a major debate in the Cognitive Science community over whether or it is is premature to offer PhDs in Cognitive Science. There are no departments yet (so no jobs in academia) and most industry has not heard of the concept. (There are some exceptions in industry, mostly the major research labs (Xerox PARC, IBM, MCC, Bell Labs, Bellcore). UCSD is considering starting a department. The Dean is establishing a high powered committee to look over the program and make recommendations. It would take from 2 to 5 years to get a department. (Establishing a new department in a new field is a major undertaking. And it requires approval of umpteen campus committees, umpteen state-wide committees, a state overseeing body for higher education in general, the regents, and probably the US senate.) I would appreciate hearing updates from people associated with other programs/departments/groups in Cognitive Science. Donald A. Norman Institute for Cognitive Science C-015 University of California, San Diego La Jolla, California 92093 norman@nprdc.arpa norman@ics.ucsd.EDU ------------------------------ Date: Thu 6 Nov 1986 10:17:14 From: ether.allegra%btl.csnet@RELAY.CS.NET Subject: non-monotonic reasoning John Nagle, in a recent posting, writes: > Non-monotonic reasoning is an attempt to make reasoning systems > less brittle, by containing the damage that can be caused by > contradiction in the axioms. The rules of inference of non-monotonic > reasoning systems are weaker than those of traditional logic. Most nonmonotonic reasoning formalisms I know of (default logic, autoepistemic logic, circumscription, NML I and II, ...) incorporate a first-order logic as a subset. Their rules of inference are thus *stronger* than traditional logics'. I think Nagle is thinking of Relevance Logic (see Anderson & Belnap), which does make an effort to contain the effects of contradiction by weakening the inference rules (avoiding the paradoxes of implication). As for truth-maintenance systems, contrary to Nagle and popular mythology, these systems typically do *not* avoid contradictions per se. What they *do* do is prevent one from 'believing' all of a set of facts explicitly marked as contradictory by the system using the TMS. These systems don't usually have any deductive power at all, they are merely constraint satisfaction devices. David W. Etherington AT&T Bell Laboratories 600 Mountain Avenue Murray Hill, NJ 07974-2070 ether%allegra@btl.csnet ------------------------------ Date: 7 Nov 86 23:06:28 GMT From: voder!lewey!evp@ucbvax.Berkeley.EDU (Ed Post) Subject: Re: Canonical list of sentient computer novels > Xref: lewey net.sf-lovers:5135 net.ai:549 > > > > I am trying to compile a canonical list of SF *novels* dealing with (1) > sentient computers, and (2) human mental access to computers or computer > networks..... Some of the classics: RUR (Rossum's Universal Robots), Carel Capek(?) Asimov's entire robot series When Harlie was One, David Gerrold The Moon is a Harsh Mistress, Robert Heinlein Colossus (sp?), The Forbin Project -- Ed Post {hplabs,voder,pyramid}!lewey!evp American Information Technology 10201 Torre Ave. Cupertino CA 95014 (408)252-8713 ------------------------------ Date: Thu, 6 Nov 86 09:40 EST From: Tim Finin Subject: AI in Literature AIList used to include frequent notes about how AI was being presented in literature, movies and TV shows. I just ran across a new wrinkle. My daughter recently bought several paperbacks published by New American Library (Signet) in a series called "The AI Gang". Here is the text from the jacket of the first book in the series, "Operation Sherlock": "Five whiz kids who call themselves the AI gang -- for Artificial Intelligence -- accompany their scientist parents to a small secluded island. While their parents are teaching a secret computer to think for itself, the kids try their hand at programming a sleuthing computer named Sherlock. They soon discover that there is an evil spy out to destroy their parents' project. When three of the gang are almost killed in an explosion, the kids and their specially developed crime computer must race against time to reveal the spy's identity ... before all of them are blown to smithereens ..." My daughter thought all of the books in the series were pretty good, btw. Tim ------------------------------ End of AIList Digest ********************