Sections 3.16 and 3.23 of Roger Penrose's Shadows of the mind (Oxford, Oxford University Press, 1994) contain a subtle and intriguing new argument against mechanism, the thesis that the human mind can be accurately modeled by a Turing machine. The argument, based on the incompleteness theorem, is designed to meet standard objections to the original Lucas-Penrose formulations. The new argument, however, seems to invoke an unrestricted truth predicate (and an unrestricted knowability predicate). If so, its premises are inconsistent. The usual (...) ways of restricting the predicates either invalidate Penrose's reasoning or require presuppositions that the mechanist can reject. (shrink)

Artificial Neural Networks have reached “grandmaster” and even “super-human” performance across a variety of games, from those involving perfect information, such as Go, to those involving imperfect information, such as “Starcraft”. Such technological developments from artificial intelligence (AI) labs have ushered concomitant applications across the world of business, where an “AI” brand-tag is quickly becoming ubiquitous. A corollary of such widespread commercial deployment is that when AI gets things wrong—an autonomous vehicle crashes, a chatbot exhibits “racist” behavior, automated credit-scoring processes (...) “discriminate” on gender, etc.—there are often significant financial, legal, and brand consequences, and the incident becomes major news. As Judea Pearl sees it, the underlying reason for such mistakes is that “... all the impressive achievements of deep learning amount to just curve fitting.” The key, as Pearl suggests, is to replace “reasoning by association” with “causal reasoning” —the ability to infer causes from observed phenomena. It is a point that was echoed by Gary Marcus and Ernest Davis in a recent piece for theNew York Times: “we need to stop building computer systems that merely get better and better at detecting statistical patterns in data sets—often using an approach known as ‘Deep Learning’—and start building computer systems that from the moment of their assembly innately grasp three basic concepts: time, space, and causality.” In this paper, foregrounding what in 1949 Gilbert Ryle termed “a category mistake”, I will offer an alternative explanation for AI errors; it is not so much that AI machinery cannot “grasp” causality, but that AI machinery (qua computation) cannot understand anything at all. (shrink)

Certain selected issues around the Gödelian anti-mechanist arguments which have received less attention are discussed.

A talk delivered at the conference “Science and Religion: The Religious Beliefs and Practices of Scientists—20th Century,” Institut für Wissenschaftsgeschichte, Georg-August-Universität, Göttingen, 28 May 2002.

More intellectual modesty, but also conceptual clarity is urgently needed in AI, perhaps more than in many other disciplines. AI research has been coined by hypes and hubris since its early beginnings in the 1950s. For instance, the Nobel laureate Herbert Simon predicted after his participation in the Dartmouth workshop that “machines will be capable, within 20 years, of doing any work that a man can do”. And expectations are in some circles still high to overblown today. This paper addresses (...) the demand for conceptual clarity and introduces precise definitions of “strong AI”, “superintelligence”, the “technological singularity”, and “artificial general intelligence” which ground in the work by the computer scientist Judea Pearl and the psychologist Howard Gardner. These clarifications allow us to embed famous arguments from the philosophy of AI in a more analytic context. (shrink)

In his PhD thesis (1938) Turing introduced what he described as 'a new kind of machine'. He called these 'O-machines'. The present paper employs Turing's concept against a number of currently fashionable positions in the philosophy of mind.

We argue that thinking of the man-machine comparison in terms of a contest involves, in a reasonable scenario, a criterion of success that is neutral. This is because we want to avoid a petitio principii. We submit, however, that, by looking at things this way, one makes the most essential human things invisible. Thus, in a sense, the contest approach is self-defeating.

Having, as it is generally agreed, failed to destroy the computational conception of mind with the G\"{o}delian attack he articulated in his {\em The Emperor's New Mind}, Penrose has returned, armed with a more elaborate and more fastidious G\"{o}delian case, expressed in and 3 of his {\em Shadows of the Mind}. The core argument in these chapters is enthymematic, and when formalized, a remarkable number of technical glitches come to light. Over and above these defects, the argument, at best, is (...) an instance of either the fallacy of denying the antecedent, the fallacy of {\em petitio principii}, or the fallacy of equivocation. More recently, writing in response to his critics in the electronic journal {\em Psyche}, Penrose has offered a G\"{o}delian case designed to improve on the version presented in {\em SOTM}. But this version is yet again another failure. In falling prey to the errors we uncover, Penrose's new G\"{o}delian case is unmasked as the same confused refrain J.R. Lucas initiated 35 years ago. (shrink)

Штучний інтелект — це комп’ютерна система, яка мислить або діє як людина. Особливості ШІ втілюють імпліцитні уявлення його розробників про природу людини. «Сильний» ШІ, який має загальні коґнітивні здатності дорослої людини, досі не створено, натомість «слабкі» ШІ вже є частиною планетарної обчислювальної інфраструктури. Нейронно-мережеві ШІ імі- тують специфічні типи людської поведінки, узагальнюючи дані про повсякденне життя ко- ристувачів. Такий підхід у створенні ШІ відповідає філософському мейнстріму ХХ сторіччя, коли повсякденне життя розглядали як джерело мовних та соціальних наперед-даностей, що забезпечують порозуміння. (...) Також цей підхід спирається на класичну дихотомію «людина— машина» й відповідне уявлення, що людська природа є стабільною та незалежною від техно- логічного стану. Проте у постметафізичну добу, коли взаємодія людини з технікою є комуні- кативною, а не інструментальною, дані про людський побут не можуть бути незалежним зразком людськості. ШІ-застосунки не лише кодифікують дескриптивні аспекти людської природи, а й дисциплінують своїх користувачів, адже цифрове середовище, в якому можуть бути зібрані дані про повсякдення, уже впорядковане за допомогою ШІ. Відповідно, у цифро- во му середовищі люди змушені відтворювати нові норми поведінки, кодифіковані за допомо- гою ШІ, який є одним із способів видозміни людської природи самою людиною. Вплив ШІ непомітний, позаяк цифрове середовище, в якому люди взаємодіють з ШІ, організовано не- достат ньо зрозумілим для користувачів чином. Антропотехнологічний характер ШІ є по бічним наслідком розвитку платформ, тому творці ШІ не перебирають на себе відповідальність за вті лені в їхніх розробках норми. (shrink)

It has been argued, by Penrose and others, that Gödel's proof of his first incompleteness theorem shows that human mathematics cannot be captured by a formal system F: the Gödel sentence G(F) of F can be proved by a (human) mathematician but is not provable in F. To this argment it has been objected that the mathematician can prove G(F) only if (s)he can prove that F is consistent, which is unlikely if F is complicated. Penrose has invented a new (...) argument intended to avoid this objection. In the paper I try to show that Penrose's new argument is inconclusive. (shrink)

The reductionist/holist debate is highly polarised. I propose an intermediate position of pragmatic holism. It derives from two claims: firstly, that irrespective of whether all natural systems are theoretically reducible, for many systems it is utterly impractical to attempt such a reduction, and secondly, that regardless of whether irreducible 'wholes exist, it is vain to try and prove this. This position illuminates the debate along new pragmatic lines by refocussing attention on the underlying heuristics of learning about the natural world.

This paper presents, from a historical and logical-philosophical perspective, the Gödelian arguments of two Oxford scholars, John Lucas and Roger Penrose. Both have been based on Gödel's Theorem to refute mechanism, computationalism and the possibility of creating an AI capable of simulating or duplicating the human mind. In the conclusions, the growing application of empirical methods in mathematics is mentioned and a possible path that would support Lucas and Penrose's arguments is speculated.

The reductionist/holist debate seems an impoverished one, with many participants appearing to adopt a position first and constructing rationalisations second. Here I propose an intermediate position of pragmatic holism, that irrespective of whether all natural systems are theoretically reducible, for many systems it is completely impractical to attempt such a reduction, also that regardless if whether irreducible `wholes' exist, it is vain to try and prove this in absolute terms. This position thus illuminates the debate along new pragmatic lines, and (...) refocusses attention on the underlying heuristics of learning about the natural world. (shrink)