I propose to consider the question, "Can machines think?" This should begin with definitions of the meaning of the terms "machine" and "think." The definitions might be framed so as to reflect so far as possible the normal use of the words, but this attitude is dangerous, If the meaning of the words "machine" and "think" are to be found by examining how they are commonly used it is difficult to escape the conclusion that the meaning and the answer (...) to the question, "Can machines think?" is to be sought in a statistical survey such as a Gallup poll. But this is absurd. Instead of attempting such a definition I shall replace the question by another, which is closely related to it and is expressed in relatively unambiguous words. The new form of the problem can be described in terms of a game which we call the 'imitation game." It is played with three people, a man (A), a woman (B), and an interrogator (C) who may be of either sex. The interrogator stays in a room apart front the other two. The object of the game for the interrogator is to determine which of the other two is the man and which is the woman. He knows them by labels X and Y, and at the end of the game he says either "X is A and Y is B" or "X is B and Y is A." The interrogator is allowed to put questions to A and B. We now ask the question, "What will happen when a machine takes the part of A in this game?" Will the interrogator decide wrongly as often when the game is played like this as he does when the game is played between a man and a woman? These questions replace our original, "Can machines think?". (shrink)

Though less well known than his other work, Turings 1938 Princeton Thesis, this title which includes his notion of an oracle machine, has had a lasting influence on computer science and mathematics. It presents a facsimile of the original typescript of the thesis along with essays by Appel and Feferman that explain its still-unfolding significance.

The classical view of computing positions computation as a closed-box transformation of inputs (rational numbers or finite strings) to outputs. According to the interactive view of computing, computation is an ongoing interactive process rather than a function-based transformation of an input to an output. Specifically, communication with the outside world happens during the computation, not before or after it. This approach radically changes our understanding of what is computation and how it is modeled. The acceptance of interaction as a new (...) paradigm is hindered by the Strong Church–Turing Thesis (SCT), the widespread belief that Turing Machines (TMs) capture all computation, so models of computation more expressive than TMs are impossible. In this paper, we show that SCT reinterprets the original Church–Turing Thesis (CTT) in a way that Turing never intended; its commonly assumed equivalence to the original is a myth. We identify and analyze the historical reasons for the widespread belief in SCT. Only by accepting that it is false can we begin to adopt interaction as an alternative paradigm of computation. We present Persistent Turing Machines (PTMs), that extend TMs to capture sequential interaction. PTMs allow us to formulate the Sequential Interaction Thesis, going beyond the expressiveness of TMs and of the CTT. The paradigm shift to interaction provides an alternative understanding of the nature of computing that better reflects the services provided by today’s computing technology. (shrink)

The classical view of computing positions computation as a closed-box transformation of inputs (rational numbers or finite strings) to outputs. According to the interactive view of computing, computation is an ongoing interactive process rather than a function-based transformation of an input to an output. Specifically, communication with the outside world happens during the computation, not before or after it. This approach radically changes our understanding of what is computation and how it is modeled. The acceptance of interaction as a new (...) paradigm is hindered by the Strong Church–Turing Thesis (SCT), the widespread belief that Turing Machines (TMs) capture all computation, so models of computation more expressive than TMs are impossible. In this paper, we show that SCT reinterprets the original Church–Turing Thesis (CTT) in a way that Turing never intended; its commonly assumed equivalence to the original is a myth. We identify and analyze the historical reasons for the widespread belief in SCT. Only by accepting that it is false can we begin to adopt interaction as an alternative paradigm of computation. We present Persistent Turing Machines (PTMs), that extend TMs to capture sequential interaction. PTMs allow us to formulate the Sequential Interaction Thesis, going beyond the expressiveness of TMs and of the CTT. The paradigm shift to interaction provides an alternative understanding of the nature of computing that better reflects the services provided by today’s computing technology. (shrink)

Since the middle of the 20th century there has been a significant debate about the attribution of capacities of living systems, particularly humans, to technological artefacts, especially computers—from Turing’s opening gambit, to subsequent considerations of artificial intelligence, to recent claims about artificial life. Some now argue that the capacities of future technologies will ultimately make it impossible to draw any meaningful distinctions between humans and machines. Such issues center on what sense, if any, it makes to claim that (...) gadgets can actually think, feel, act, live, etc. I outline this debate and offer a critique of its persistent polarization. I characterize two of the debate’s major camps (associated roughly with Turing and Searle); argue that the debate’s structure (including key assumptions inherent to each camp) precludes resolution; and, contend that some central clashes within the debate actually stem from an inadequately drawn distinction between claims about the capacities of artifacts and claims about the proper criteria for assessing such attributions. I offer a different perspective in which I: challenge some central elements of the debate that contribute to its perennially irresolvable state; hold that the debate needs to be placed more squarely in sync with how we in fact treat the attribution of such capacities to humans themselves; and, offer (unlike the other two camps) a foothold for making moral assessments of such proposed technologies. (shrink)

Since the middle of the 20th century there has been a significant debate about the attribution of capacities of living systems, particularly humans, to technological artefacts, especially computers—from Turing’s opening gambit, to subsequent considerations of artificial intelligence, to recent claims about artificial life. Some now argue that the capacities of future technologies will ultimately make it impossible to draw any meaningful distinctions between humans and machines. Such issues center on what sense, if any, it makes to claim that (...) gadgets can actually think, feel, act, live, etc. I outline this debate and offer a critique of its persistent polarization. I characterize two of the debate’s major camps (associated roughly with Turing and Searle); argue that the debate’s structure (including key assumptions inherent to each camp) precludes resolution; and, contend that some central clashes within the debate actually stem from an inadequately drawn distinction between claims about the capacities of artifacts and claims about the proper criteria for assessing such attributions. I offer a different perspective in which I: challenge some central elements of the debate that contribute to its perennially irresolvable state; hold that the debate needs to be placed more squarely in sync with how we in fact treat the attribution of such capacities to humans themselves; and, offer (unlike the other two camps) a foothold for making moral assessments of such proposed technologies. (shrink)

Accelerating Turing machines are Turing machines of a sort able to perform tasks that are commonly regarded as impossible for Turing machines. For example, they can determine whether or not the decimal representation of contains n consecutive 7s, for any n; solve the Turing-machine halting problem; and decide the predicate calculus. Are accelerating Turing machines, then, logically impossible devices? I argue that they are not. There are implications concerning the nature of (...) effective procedures and the theoretical limits of computability. Contrary to a recent paper by Bringsjord, Bello and Ferrucci, however, the concept of an accelerating Turing machine cannot be used to shove up Searle's Chinese room argument. (shrink)

Accelerated Turing machines are Turing machines that perform tasks commonly regarded as impossible, such as computing the halting function. The existence of these notional machines has obvious implications concerning the theoretical limits of computability.

In a recent study of a patient in a persistent vegetative state, [Owen, A. M., Coleman, M. R., Boly, M., Davis, M. H., Laureys, S., & Pickard, J. D. . Detecting awareness in the vegetative state. Science, 313, 1402] claimed that they had demonstrated the presence of consciousness in this patient. This bold conclusion was based on the isomorphy between brain activity in this patient and a set of conscious control subjects, obtained in various imagery tasks. However, establishing consciousness (...) in unresponsive patients is fraught with methodological and conceptual difficulties. The aim of this paper is to demonstrate that the current debate surrounding consciousness in VS patients has parallels in the artificial intelligence debate as to whether machines can think. Basically, used a method analogous to the Turing test to reveal the presence of consciousness, whereas their adversaries adopted a line of reasoning akin to Searle’s Chinese room argument. Highlighting the correspondence between these two debates can help to clarify the issues surrounding consciousness in non-communicative agents. (shrink)

The tape is divided into squares, each square bearing a single symbol—'0' or '1', for example. This tape is the machine's general-purpose storage medium: the machine is set in motion with its input inscribed on the tape, output is written onto the tape by the head, and the tape serves as a short-term working memory for the results of intermediate steps of the computation. The program governing the particular computation that the machine is to perform is also stored on the (...) tape. A small, fixed program that is 'hard-wired' into the head enables the head to read and execute the instructions of whatever program is on the tape. The machine's atomic operations are very simple—for example, 'move left one square', 'move right one square', 'identify the symbol currently beneath the head', 'write 1 on the square that is beneath the head', and 'write 0 on the square that is beneath the head'. Complexity of operation is achieved by the chaining together of large numbers of these simple atoms. Any universal Turing machine can be programmed to carry out any calculation that can be performed by a human mathematician working with paper and pencil in accordance with some algorithmic method. This is what is meant by calling these machines 'universal'. (shrink)

Since the beginning of the twenty-first century there has been an increasing awareness that software rep- resents a blind spot in new media theory. The growing interest in software also influences the argument in this paper, which sets out from the assumption that Alan M. Turing's concept of the universal machine, the first theoretical description of a computer program, is a kind of bachelor machine. Previous writings based on a similar hypothesis have focused either on a comparison of the (...) universal machine and the bachelor machine in terms of the similarities of their structural features, or they have taken the bachelor machine as a metaphor for a man or a computer. Unlike them, this paper stresses the importance of the con- text as a key to interpreting the universal Turing machine as a bachelor machine and, potentially, as a self-portrait. (shrink)

Accelerating Turing machines have attracted much attention in the last decade or so. They have been described as “the work-horse of hypercomputation” (Potgieter and Rosinger 2010: 853). But do they really compute beyond the “Turing limit”—e.g., compute the halting function? We argue that the answer depends on what you mean by an accelerating Turing machine, on what you mean by computation, and even on what you mean by a Turing machine. We show first that in (...) the current literature the term “accelerating Turing machine” is used to refer to two very different species of accelerating machine, which we call end-stage-in and end-stage-out machines, respectively. We argue that end-stage-in accelerating machines are not Turing machines at all. We then present two differing conceptions of computation, the internal and the external, and introduce the notion of an epistemic embedding of a computation. We argue that no accelerating Turing machine computes the halting function in the internal sense. Finally, we distinguish between two very different conceptions of the Turing machine, the purist conception and the realist conception; and we argue that Turing himself was no subscriber to the purist conception. We conclude that under the realist conception, but not under the purist conception, an accelerating Turing machine is able to compute the halting function in the external sense. We adopt a relatively informal approach throughout, since we take the key issues to be philosophical rather than mathematical. (shrink)

Storrs McCall appeals to a particular true but improvable sentence of formal arithmetic to argue, by appeal to its irrefutability, that human minds transcend Turing machines. Metamathematical oversights in McCall's discussion of the Godel phenomena, however, render invalid his philosophical argument for this transcendentalist conclusion.

In I used Turing machine arguments to show that computers can recognize humanly recognizable patterns in principle. In 1978 James D. Heffernan has expressed some doubts about such arguments. He does not question the propositions that I defend in the paper, nor the specific arguments in their support. What he does criticize are certain background assumptions.

Infinite time Turing machines with only one tape are in many respects fully as powerful as their multi-tape cousins. In particular, the two models of machine give rise to the same class of decidable sets, the same degree structure and, at least for partial functions f : ℝ → ℕ, the same class of computable functions. Nevertheless, there are infinite time computable functions f : ℝ → ℝ that are not one-tape computable, and so the two models of (...) infinitary computation are not equivalent. Surprisingly, the class of one-tape computable functions is not closed under composition; but closing it under composition yields the full class of all infinite time computable functions. Finally, every ordinal that is clockable by an infinite time Turing machine is clockable by a one-tape machine, except certain isolated ordinals that end gaps in the clockable ordinals. (shrink)

Infinite time Turing machines extend the operation of ordinary Turing machines into transfinite ordinal time. By doing so, they provide a natural model of infinitary computability, a theoretical setting for the analysis of the power and limitations of supertask algorithms.

Infinite time Turing machines extend the operation of ordinary Turing machines into transfinite ordinal time. By doing so, they provide a natural model of infinitary computability, a theoretical setting for the analysis of the power and limitations of supertask algorithms.

Can mind be modeled as a Turing machine? If you find such questions irrelevant, e.g. because the subject is already exhausted, then you need not read the book Mind versus Computer (Gams et al., 1991). If, on the other hand, you do find such questions relevant, then perhaps you need not read Dunlop's review of the book (Dunlop, 2000). (...).

We describe an emerging field, that of nonclassical computability and nonclassical computing machinery. According to the nonclassicist, the set of well-defined computations is not exhausted by the computations that can be carried out by a Turing machine. We provide an overview of the field and a philosophical defence of its foundations.

We extend in a natural way the operation of Turing machines to infinite ordinal time, and investigate the resulting supertask theory of computability and decidability on the reals. Everyset. for example, is decidable by such machines, and the semi-decidable sets form a portion of thesets. Our oracle concept leads to a notion of relative computability for sets of reals and a rich degree structure, stratified by two natural jump operators.

We first discuss Michael Dummett’s philosophy of mathematics and Robert Brandom’s philosophy of language to demonstrate that inferentialism entails the falsity of Church’s Thesis and, as a consequence, the Computational Theory of Mind. This amounts to an entirely novel critique of mechanism in the philosophy of mind, one we show to have tremendous advantages over the traditional Lucas-Penrose argument.

We characterise explicitly the decidable predicates on integers of Infinite Time Turing machines, in terms of admissibility theory and the constructible hierarchy. We do this by pinning down ζ, the least ordinal not the length of any eventual output of an Infinite Time Turing machine (halting or otherwise); using this the Infinite Time Turing Degrees are considered, and it is shown how the jump operator coincides with the production of mastercodes for the constructible hierarchy; further that (...) the natural ordinals associated with the jump operator satisfy a Spector criterion, and correspond to the L ζ -stables. It also implies that the machines devised are "Σ 2 Complete" amongst all such other possible machines. It is shown that least upper bounds of an "eventual jump" hierarchy exist on an initial segment. (shrink)

Vann McGee has argued that, given certain background assumptions and an ought-implies-can thesis about norms of rationality, Bayesianism conflicts globally with computationalism due to the fact that Robinson arithmetic is essentially undecidable. I show how to sharpen McGee's result using an additional fact from recursion theory—the existence of a computable sequence of computable reals with an uncomputable limit. In conjunction with the countable additivity requirement on probabilities, such a sequence can be used to construct a specific proposition to which Bayesianism (...) requires an agent to assign uncomputable credence—yielding a local conflict with computationalism. (shrink)

The common view that the notion of a Turing machine is directly relevant to AI is criticised. It is argued that computers are the result of a convergence of two strands of development with a long history: development of machines for automating various physical processes and machines for performing abstract operations on abstract entities, e.g. doing numerical calculations. Various aspects of these developments are analysed, along with their relevance to AI, and the similarities between computers viewed in (...) this way and animal brains. This comparison depends on a number of distinctions: between energy requirements and information requirements of machines, between ballistic and online control, between internal and external operations, and between various kinds of autonomy and self-awareness. The ideas are all intuitively familiar to software engineers, though rarely made fully explicit. Most of this has nothing to do with Turing machines or most of the mathematical theory of computation. But it has everything to do with both the scientific task of understanding, modelling or replicating human or animal intelligence and the engineering applications of AI, as well as other applications of computers. (shrink)

The properties of Turing’s famous ‘universal machine’ has long sustained functionalist intuitions about the nature of cognition. Here, I show that there is a logical problem with standard functionalist arguments for multiple realizability. These arguments rely essentially on Turing’s powerful insights regarding computation. In addressing a possible reply to this criticism, I further argue that functionalism is not a useful approach for understanding what it is to have a mind. In particular, I show that the difficulties involved in (...) distinguishing implementation from function make multiple realizability claims untestable and uninformative. As a result, I conclude that the role of Turing machines in philosophy of mind needs to be reconsidered. (shrink)

Earlier, we have studied computations possible by physical systems and by algorithms combined with physical systems. In particular, we have analysed the idea of using an experiment as an oracle to an abstract computational device, such as the Turing machine. The theory of composite machines of this kind can be used to understand (a) a Turing machine receiving extra computational power from a physical process, or (b) an experimenter modelled as a Turing machine performing a test (...) of a known physical theory T.Our earlier work was based upon experiments in Newtonian mechanics. Here we extend the scope of the theory of experimental oracles beyond Newtonian mechanics to electrical theory. First, we specify an experiment that measures resistance using a Wheatstone bridge and start to classify the computational power of this experimental oracle using non-uniform complexity classes. Secondly, we show that modelling an experimenter and experimental procedure algorithmically imposes a limit on our ability to measure resistance by the Wheatstone bridge.The connection between the algorithm and physical test is mediated by a protocol controlling each query, especially the physical time taken by the experimenter. In our studies we find that physical experiments have an exponential time protocol; this we formulate as a general conjecture. Our theory proposes that measurability in Physics is subject to laws which are co-lateral effects of the limits of computability and computational complexity. (shrink)

Earlier, we have studied computations possible by physical systems and by algorithms combined with physical systems. In particular, we have analysed the idea of using an experiment as an oracle to an abstract computational device, such as the Turing machine. The theory of composite machines of this kind can be used to understand (a) a Turing machine receiving extra computational power from a physical process, or (b) an experimenter modelled as a Turing machine performing a test (...) of a known physical theory T. Our earlier work was based upon experiments in Newtonian mechanics. Here we extend the scope of the theory of experimental oracles beyond Newtonian mechanics to electrical theory. First, we specify an experiment that measures resistance using a Wheatstone bridge and start to classify the computational power of this experimental oracle using non-uniform complexity classes. Secondly, we show that modelling an experimenter and experimental procedure algorithmically imposes a limit on our ability to measure resistance by the Wheatstone bridge. The connection between the algorithm and physical test is mediated by a protocol controlling each query, especially the physical time taken by the experimenter. In our studies we find that physical experiments have an exponential time protocol, this we formulate as a general conjecture. Our theory proposes that measurability in Physics is subject to laws which are co-lateral effects of the limits of computability and computational complexity. (shrink)

We characterise explicitly the decidable predicates on integers of Infinite Time Turing machines, in terms of admissibility theory and the constructible hierarchy. We do this by pinning down $\zeta$, the least ordinal not the length of any eventual output of an Infinite Time Turing machine ; using this the Infinite Time Turing Degrees are considered, and it is shown how the jump operator coincides with the production of mastercodes for the constructible hierarchy; further that the natural (...) ordinals associated with the jump operator satisfy a Spector criterion, and correspond to the L$_\zeta$-stables. It also implies that the machines devised are "$\Sigma_2$ Complete" amongst all such other possible machines. It is shown that least upper bounds of an "eventual jump" hierarchy exist on an initial segment. (shrink)

Infinite time Turing machines extend the operation of ordinary Turing machines into transfinite ordinal time. By doing so, they provide a natural model of infinitary computability, a theoretical setting for the analysis of the power and limitations of supertask algorithms.

In his article “On Mechanical Recognition” R. J. Nelson brings to bear a branch of mathematical logic called automata theory on problems of artificial intelligence. Specifically he attacks the anti-mechanist claim that “[i]nasmuch as human recognition to a very great extent relies on context and on the ability to grasp wholes with some independence of the quality of the parts, even to fill in the missing parts on the basis of expectations, it follows that computers cannot in principle be programmed (...) to recognize or learn to recognize all patterns”, p. 24). Nelson proposes, contrary to this claim, that “gestalt recognition is not beyond digital automata”. in particular, he claims, he will establish by what he calls “Turing machine arguments” that the following four theses are true: automata can recognize different pattern types in one and the same set of instances;automata can recognize the “same” pattern in different sets of instances;automata can recognize incomplete, degraded patterns having missing or indeterminate parts;automata can recognize family resemblances. (shrink)

Church's thesis asserts that a number-theoretic function is intuitively computable if and only if it is recursive. A related thesis asserts that Turing's work yields a conceptual analysis of the intuitive notion of numerical computability. I endorse Church's thesis, but I argue against the related thesis. I argue that purported conceptual analyses based upon Turing's work involve a subtle but persistent circularity. Turing machines manipulate syntactic entities. To specify which number-theoretic function a Turing machine (...) computes, we must correlate these syntactic entities with numbers. I argue that, in providing this correlation, we must demand that the correlation itself be computable. Otherwise, the Turing machine will compute uncomputable functions. But if we presuppose the intuitive notion of a computable relation between syntactic entities and numbers, then our analysis of computability is circular. (shrink)

Lucas and Penrose have contended that, by displaying how any characterisation of arithmetical proof programmable into a machine allows of diagonalisation, generating a humanly recognisable proof which eludes that characterisation, Gödel's incompleteness theorem rules out any purely mechanical model of the human intellect. The main criticisms of this argument have been that the proof generated by diagonalisation (i) will not be humanly recognisable unless humans can grasp the specification of the object-system (Benacerraf); and (ii) counts as a proof only on (...) the (unproven) hypothesis that the object system is consistent (Putnam). The present paper argues that criticism (ii) may be met head-on by an intuitionistic proponent of the anti-mechanist argument; and that criticism (i) is simply mistaken. However the paper concludes by questioning the sufficiency of the situation for an interesting anti-mechanist conclusion. (shrink)