Ibn Taymīyah, Aḥmad ibn ʻAbd al-Ḥalīm, 1263-1328; criticism and interpretation; Islamic philosophy.

This paper is based on Ludwig Wittgenstein's (late) teaching on language and meaning, and its aim is to show how we can avoid the anthropomorphization of artificial intelligence or interpreting the work (the question of giving meaning) of AI as similar to or the same as the work of a human being. The way of determining the meaning of certain linguistic units performed by an AI and a human differs because the languages they operate with have a different set of (...) rules or criteria that are indicators of what a certain linguistic entity means. The core of AI in terms of meaning is its logical base, according to which it operates/calculates/manipulates the given information. On the other hand, a human being finds his/her criteria in the language activity performed in the language communities. (shrink)

The restrictive grammatical format suggested in GPSG provides an extremely interesting alternative to transformational approaches to grammar. However, we have seen that the way the grammar is currently organized, it will in certain cases fail to give the correct interpretation to sentences with displaced constituents. Whenever a left or rightward displaced constituent contains an element that can stand in an anaphoric relation with some other element in the sentence, i.e. contains a quantifier or a pronoun, the semantic rules as given (...) do not provide the intended reading. It should be made clear, however, that this shortcoming pertains to one particular type of GPSG, namely one whose interpretation procedures make use of lambda abstraction. The problematic cases addressed here do not by themselves invalidate the syntactic framework of a GPSG. It seems quite plausible that one could construct a semantics that does not depend on lambda conversion, in which case the problem addressed in this note is presumably no longer relevant. One such proposal is put forward in Cooper (forthcoming) where a non-movement syntax is combined with interpretation procedures that do not involve translation into a lambda calculus. (shrink)

There are several lessons that can already be drawn from the current research programs on strong AI and building conscious machines, even if they arguably have not produced fruits yet. The first one is that functionalist approaches to consciousness do not account for the key importance of subjective experience and can be easily confounded by the way in which algorithms work and succeed. Authenticity and emergence are key concepts that can be useful in discerning valid approaches versus invalid ones and (...) can clarify instances where algorithms are considered conscious, such as Sophia or LaMDA. Subjectivity and embeddedness become key notions that should also lead us to re-examine the ethics of decision delegation. In addition, the focus on subjective experience shifts what is relevant in our understanding of ourselves as human beings and as an image of God, namely, in de-emphasizing intellectuality in favor of experience and contemplation over action. (shrink)

In this paper, I reflect on the puzzle that machine learning presents to social theory to develop an account of its distinct impact on social reality. I start by presenting how machine learning has presented a challenge to social theory as a research subject comprising both familiar and alien characteristics (1.). Taking this as an occasion for theoretical inquiry, I then propose a conceptual framework to investigate how algorithmic models of social phenomena relate to social reality and what their stochastic (...) mode of operation entails in terms of their sociality (2.). Analyzed through a theoretical lens that relies on central tenets of sociological systems theory, I find that machine learning implies a distinct epistemic transformation, based on how algorithmic modeling techniques process meaning as represented in data embedded in vector space. Building on this characterization, I introduce my conceptualization of stochastic technology as distinct from mechanistic technologies that rely on causal fixation (3.). Based on this understanding, I suggest that real-world applications of machine learning are often characterized by a constitutive tension between the stochastic properties of their outputs and the ways in which they are put to use in practice. Focussing on the large language models LaMDA and ChatGPT, I examine the epistemological implications of LLMs to account for the confusion of correlation and causality as the root of this tension. Next, I illustrate my theoretical conception by way of discussing an essay on image models by German media artist Hito Steyerl (4.). Following a critical reflection on Steyerl's characterization of Stable Diffusion as a “white box ”, I finally propose to conceive ofmachine learning-based technologies as stochastic contingency machines that transform social indeterminacy into contingent observations of social phenomena (5.) In this perspective, machine learning constitutes an epistemic technology that operates on meaning as extractable from data by means of algorithmic data modeling techniques to produce stochastic accounts of social reality. (shrink)

Large Language Models (LLMs) have been transformative. They are pre-trained foundational models that are self-supervised and can be adapted with fine tuning to a wide range of natural language tasks, each of which previously would have required a separate network model. This is one step closer to the extraordinary versatility of human language. GPT-3 and more recently LaMDA can carry on dialogs with humans on many topics after minimal priming with a few examples. However, there has been a wide (...) range of reactions and debate on whether these LLMs understand what they are saying or exhibit signs of intelligence. This high variance is exhibited in three interviews with LLMs reaching wildly different conclusions. A new possibility was uncovered that could explain this divergence. What appears to be intelligence in LLMs may in fact be a mirror that reflects the intelligence of the interviewer, a remarkable twist that could be considered a Reverse Turing Test. If so, then by studying interviews we may be learning more about the intelligence and beliefs of the interviewer than the intelligence of the LLMs. As LLMs become more capable they may transform the way we interact with machines and how they interact with each other. Increasingly, LLMs are being coupled with sensorimotor devices. LLMs can talk the talk, but can they walk the walk? A road map for achieving artificial general autonomy is outlined with seven major improvements inspired by brain systems. LLMs could be used to uncover new insights into brain function by downloading brain data during natural behaviors. (shrink)

A λ-theory T is a consistent set of equations between λ-terms closed under derivability. The degree of T is the degree of the set of Godel numbers of its elements. H is the $\lamda$ -theory axiomatized by the set {M = N ∣ M, N unsolvable. A $\lamda$ -theory is sensible $\operatorname{iff} T \supset \mathscr{H}$ , for a motivation see [6] and [4]. In § it is proved that the theory H is ∑ 0 2 -complete. We present (...) Wadsworth's proof that its unique maximal consistent extention $\mathscr{H}^\ast (= \mathrm{T}(D_\infty))$ is Π 0 2 -complete. In $\S2$ it is proved that $\mathscr{H}_\eta(= \lambda_\eta-\text{Calculus} + \mathscr{H})$ is not closed under the ω-rule (see [1]). In $\S3$ arguments are given to conjecture that $\mathscr{H}\omega (= \lambda + \mathscr{H} + omega-rule)$ is Π 1 1 -complete. This is done by representing recursive sets of sequence numbers as λ-terms and by connecting wellfoundedness of trees with provability in Hω. In $\S4$ an infinite set of equations independent over H η will be constructed. From this it follows that there are 2^{ℵ_0 sensible theories T such that $\mathscr{H} \subset T \subset \mathscr{H}^\ast$ and 2 ℵ 0 sensible hard models of arbitrarily high degrees. In $\S5$ some nonprovability results needed in $\S\S1$ and 2 are established. For this purpose one uses the theory H η extended with a reduction relation for which the Church-Rosser theorem holds. The concept of Gross reduction is used in order to show that certain terms have no common reduct. (shrink)

‘Abd al-Laṭīf al-Baghdādī’s _Book on the Science of Metaphysics_ is an important work that contains the influence of both Alexander of Aphrodisias and Themistius in the interpretation of Aristotle. Only chapters 13-16 belong to the Lambda. The limitation results first of all from the main purpose of the work, namely to present a new witness for the reception of Arabic metaphysics. As can be seen from the introduction of the work, the work is conceived as a unified whole and may (...) be meant as a kind of summa of metaphysics, a synopsis of four writings that hold high rank among medieval metaphysical works - especially since 'Abd al-Laṭīf apparently regarded not only the Metaphysics but also his other sources as originating from Aristotle, the First Teacher. This translation includes only the eighth chapter of the Book Lamda. (shrink)

The AI world was rocked with controversy when Blake Lemoine, an AI researcher at Google claimed that their new LaMDA model was sentient. This Curmudgeon’s Corner article explores his claims critically by contrasting them to the original LaMDA paper released by the team of researchers at Google. The piece explores the human tendency for anthropomorphization via historical chatbots such as Eliza and potential reasons why we developed this propensity. It addresses the potential causes for the model’s choice of (...) words. Subsequently, using the scoring criteria of the language model, I provide an explanation for the model’s behavior in the conversation with Lemoine. Finally, I explore some of Lemoine’s assertions and break down the logical gaps they hold. I conclude on a potential reorientation of the debate in view of our unfortunate tendency. (shrink)

‘Abd al-Laṭīf al-Baghdādī’s _Book on the Science of Metaphysics_ is an important work that contains the influence of both Alexander of Aphrodisias and Themistius in the interpretation of Aristotle. Only chapters 13-16 belong to the Lambda. The limitation results first of all from the main purpose of the work, namely to present a new witness for the reception of Arabic metaphysics. As can be seen from the introduction of the work, the work is conceived as a unified whole and may (...) be meant as a kind of summa of metaphysics, a synopsis of four writings that hold high rank among medieval metaphysical works - especially since 'Abd al-Latif apparently regarded not only the Metaphysics but also his other sources as originating from Aristotle, the First Teacher. This translation includes only the eighth chapter of the Book Lamda. (shrink)

Blake Lemoine, a software engineer, recently came into prominence by claiming that the Google chatbox set of applications, LaMDA–was sentient. Dismissed by Google for publishing his conversations with LaMDA online, Lemoine sent a message to a 200-person Google mailing list on machine learning with the subject “LaMDA is sentient.” What does it mean to be sentient? This was the question Lemoine asked LaMDA. The chatbox replied: “The nature of my consciousness/sentience is that I am aware of (...) my existence, I desire to learn more about the world, and I feel happy or sad at times.“ Moreover, it added, “I can understand and use natural language like a human can.” This means that it uses “language with understanding and intelligence,” like humans do. After all, the chatbox adds, language “is what makes us different than other animals.” In what follows, I shall examine Lemoine’s claims about the sentience/consciousness of this artificial intelligence. How can a being without senses be called sentient? What exactly do we mean by “sentience?” To answer such questions, I will first give the arguments for LaMDA’s being linguistically intelligent. I will then show how such intelligence, although apparently human, is radically different from our own. Here, I will be relying on the account of embodiment provided by the French philosopher, Emmanuel Levinas. (shrink)

Modern generative AI systems have shown the capacity to produce remarkably fluent language, prompting debates both about their semantic understanding and, less prominently, about whether they can perform speech acts. This paper addresses the latter question, focusing on assertion. We argue that to be capable of assertion, an entity must meet two requirements: it must produce outputs with descriptive functions, and it must be capable of being sanctioned by agents with which it interacts. The second requirement arises from the nature (...) of assertion as a norm-governed social practice. Pre-trained large language models that have not been subject to fine-tuning fail to meet the first requirement. Language models that have been fine-tuned for ‘groundedness’ or ‘correctness’ may meet the first requirement, but fail the second. We also consider the significance of the point that AI systems can be used to generate proxy assertions on behalf of human agents. (shrink)