Ascribing mental qualities like beliefs, intentions and wants to a machine is sometimes correct if done conservatively and is sometimes necessary to express what is known about its state. We propose some new definitional tools for this: definitions relative to an approximate theory and second order structural definitions.

This collection of articles and associated discussion papers focuses on a problem that has attracted increasing attention from linguists and psychologists throughout the world during the past several years. Reduced to essentials, the problem is that of discovering the character of the mental capacities that make it possible for human beings to attain knowledge of their language on the basis of fragmentary and haphazard early linguistic experience. A fundamental assumption running through all of these contributions is that people possess strong (...) innate predispositions that are critical for success in this task. (shrink)

This article is inspired by recent psychological studies confirming that a child is not born a blank slate but has important innate capabilities. An important part of the ``learning'' required to deal with the three dimensional world of objects, processes, and other beings was done by evolution. Each child need not do this learning itself.

These notes discuss formalizing contexts as first class objects. The basic relationships are: ist(c,p) meaning that the proposition p is true in the context c, and value(c,p) designating the value of the term e in the context c Besides these, there are lifting formulas that relate the propositions and terms in subcontexts to possibly more general propositions and terms in the outer context. Subcontextx are often specialised with regard to time, place and terminology. Introducing contexts as formal objects will permit (...) axiomatizations in limited contexts to be expanded to transcend the original limitations. This seems necessary to provide AI programs using logic with certain capabilities that human fact representation and human reasoning possess. Fully implementing transcendence seems to require further extensions to mathematical logic, ie. beyond the nonmonotonic inference methods first invented in AI and now studied as a new domain of logic. (shrink)

In AI, consciousness of self consists in a program having certain kinds of facts about its own mental processes and state of mind. We discuss what consciousness of its own mental structures a robot will need in order to operate in the common sense world and accomplish the tasks humans will give it. It's quite a lot. Many features of human consciousness will be wanted, some will not, and some abilities not possessed by humans have already been found feasible and (...) useful in limited contexts. We give preliminary fragments of a logical language a robot can use to represent information about its own state of mind. A robot will often have to conclude that it cannot decide a question on the basis of the information in memory and therefore must seek information externally. Gödel's idea of relative consistency is used to formalize non-knowledge. Programs with the kind of consciousness discussed in this article do not yet exist, although programs with some components of it exist. Thinking about consciousness with a view to designing it provides a new approach to some of the problems of consciousness studied by philosophers. One advantage is that it focusses on the aspects of consciousness important for intelligent behavior. (shrink)

We discuss first order theories in which individual concepts are admitted as mathematical objects along with the things that reify them. This allows very straightforward formalizations of knowledge, belief, wanting, and necessity in ordinary first order logic without modal operators. Applications are given in philosophy and in artificial intelligence. We do not treat general concepts, and we do not present any full axiomatizations but rather show how various facts can be expressed.

Although living conditions have improved throughout history, protest, at least in the last few decades, seems to have increased to the point of becoming a normal phenomenon in modern societies. Contributors to this volume examine how and why this is the case and argue that although problems such as poverty, hunger, and violations of democratic rights may have been reduced in advanced Western societies, a variety of other problems and opportunities have emerged and multiplied the reasons and possibilities for protest.

This article is oriented toward the use of modality in artificial intelligence (AI). An agent must reason about what it or other agents know, believe, want, intend or owe. Referentially opaque modalities are needed and must be formalized correctly. Unfortunately, modal logics seem too limited for many important purposes. This article contains examples of uses of modality for which modal logic seems inadequate.I have no proof that modal logic is inadequate, so I hope modal logicians will take the examples as (...) challenges. (shrink)

Logical AI develops computer programs that represent what they know about the world primarily by logical formulas and decide what to do primarily by logical reasoning--including nonmonotonic logical reasoning. It is convenient to use logical sentences and terms whose meaning depends on context. The reasons for this are similar to what causes human language to use context dependent meanings. This note gives elements of some of the formalisms to which we have been led. Fuller treatments are in [McC93], [Guh91] and (...) [MB94] and the references cited in the Web page [Buv95]. The first main idea is to make contexts first class objects in the logic and use the formula ist(c,p) to assert that the proposition p is true in the context c. A second idea is to formalize how propositions true in one context transform when they are moved to different but related contexts. An ability to transcend the outermost context is needed to give computer programs the ability to reason about the totality of all they have thought about so far [McC96]. (shrink)

Logical AI involves representing knowledge of an agent’s world, its goals and the current situation by sentences in logic. The agent decides what to do by inferring that a certain action or course of action is appropriate to achieve the goals. We characterize briefly a large number of concepts that have arisen in research in logical AI. Reaching human-level AI requires programs that deal with the common sense informatic situation. This in turn requires extensions from the way logic has been (...) used in formalizing branches of mathematics and physical science. It also seems to require extensions to the logics themselves, both in the formalism for expressing knowledge and the reasoning used to reach conclusions. A large number of concepts need to be studied to achieve logical AI of human level. This article presents candidates. The references, though numerous, to articles concerning these concepts are still insuf- ficient, and I’ll be grateful for more, especially for papers available on the web. This article is available in several forms via http://www-formal.stanford.edu/jmc/conceptsai.html. (shrink)

From the AI point of view, consciousness must be regarded as a collection of interacting processes rather than the unitary object of much philosophical speculation. We ask what kinds of propositions and other entities need to be designed for consciousness to be useful to an animal or a machine. We thereby assert that human consciousness is useful to human functioning and not just and epiphenomenon. Zombies in the sense of Todd Moody's article are merely the victims of Moody's prejudices. To (...) behave like humans, zombies will need what Moody might call pseudo-consciousness, but useful pseudo-consciousness will share all the observable qualities of human consciousness including what the zombie will be able to report. Robots will require a pseudo-consciousness with many of the intellectual qualities of human consciousness but will function successfully with few if any human emotional conscious qualities if that is how we choose to build them. (shrink)

An 8 by 8 checkerboard with two diagonally opposite squares removed cannot be covered by dominoes each of which covers two rectilinearly adjacent squares. present a set theory description of the proposition and an informal proof that the covering is impossible. While no present system that I know of will accept either formal description or the proof, I claim that both should be admitted in any heavy duty set theory.

John Searle begins his ``Consciousness, Explanatory Inversion and Cognitive Science'' with " ``Ten years ago in this journal I published an article criticising what I call Strong AI, the view that for a system to have mental states it is sufficient for the system to implement the right sort of program with right inputs and outputs. Strong AI is rather easy to refute and the basic argument can be summarized in one sentence: {it a system, me for example, could implement (...) a program for understanding Chinese, for example, without understanding any Chinese at all.} This idea, when developed, became known as the Chinese Room Argument.'' " The Chinese Room Argument can be refuted in one sentence. (shrink)

internal events that happen spontaneously from external events (actions). It also treats processes, e.g. a buzzer, that do not settle down. The non-monotonic reasoning is circumscription done situation by situation.

Here's the article which was a 1964 Stanford AI Memo. After the original memo, several people offered different proofs of the theorem including Shmuel Winograd, Marvin Minsky and Dimitri Stefanyuk - none published, to my knowledge. Winograd claimed that his proof was non-creative, because it didn't use an extraneous idea like the colors of the squares. This set off a contest to see who could produce the most non-creative proof. Minsky's idea was to start with the diagonal next to an (...) excluded corner ssquare, note that 2 dominoes had to project from it to the diagonal with three squares, and from there 1 domino to the four square diagonal, etc. Coming from the other end also leaves only six of the eight squares in the long diagonal covered. Minsky's proof gets high points for non-creativity, because it is specific to the 8 by 8 board. (Using the colors it is easy to show that a Minsky style proof will work for any even sized board.). (shrink)

The aim of this article is to motivate and outline a dialogical perspective on agency that accommodates centrifugal and centripetal tendencies in current cultural theories of agency. To complement approaches that assume a high degree of integration and clarity, we emphasise the diversity of agency as it is experienced in the open-ended dialogical relationship with a particular other. While these former approaches to agency provide us with the means to examine the influence of social processes such as division of labour (...) and reproduction of community, they tend to underplay the importance of agency as it is embodied and experienced in the lives of particular people. To reflect on this aspect of agency, we will draw extensively on Bakhtin's work, which crosses boundaries between philosophy, psychology and literary criticism, and which is concerned to understand what is personal in activity. In terms of agency, his work draws our attention to the sense of responsibility and potentiality that imbues our dialogues with particular others, the aesthetics and ethics involved in dialoguing with the other, and the sense of dilemmatic choice and intonation that is involved in our dialogue with the other. (shrink)

This is the first science fiction story I have put up for the public to look at. While it was written just as a story, it partly illustrates my opinions about what household robots should be like. In my article Making Robots Conscious of their Mental States , I argued that robots should not be programmed to have..

Apology: My knowledge of of machine learning is no more recent than Tom Mitchell's book. Its chapters describe, except for inductive logic programming, programs aimed at classifying appearances.

Licklider--1960--Man-Computer Symbiosis Roberts--1970--ARPAnet Internet Engelbart--1962-1968--Mouse, linked documents Kay--1970--Dynabook Berners-Lee--late 1980s and early 90s--World Wide Web Brin and Page--1996--Google--first adequate search engine other prophets--Nelson, etc. whom I neglect undeservedly from ignorance.

AI needs many ideas that have hitherto been studied only by philosophers. This is because a robot, if it is to have human level intelligence and ability to learn from its experience, needs a general world view in which to organize facts. It turns out that many philosophical problems take new forms when thought about in terms of how to design a robot. Some approaches to philosophy are helpful and others are not.

Aristotle and the shape of modern friendships. With John McCarthy and Matthew West.