By Teresa (Tim) Christiaens (Tilburg University)

From his early 1844 Manuscripts or his later “Fragment on Machines” in the Grundrisse to his mature volumes of Capital, Karl Marx has consistently stressed the role of technology in capitalist domination. These criticisms have recently returned with the rise of Artificial Intelligence in the workplace, which not only threatens workers with mass unemployment but also risks subjugating the workers who remain in employment under tight managerial surveillance. Marx expressed similar worries about industrial technological development during the 19th century. He described how new steam-powered productions methods were transforming the world of work in favour of capital accumulation and to the detriment of worker autonomy and wellbeing. However, his critique shifts and turns across multiple lines of thought. As Sandro Mezzadra argues in Nei cantieri marxiani (2014), Marx’s oeuvre is composed of many “workshops” (cantieri), conceptual experiments that Marx develops and drops again every time he starts writing a new draft. Marx scholarship has subsequently attempted to synthesize these multiple critical experiments into a consistent critical theory of technology. Some highlight the problem of alienation as key to Marx’ concerns about capitalist technology (Marcuse, Feenberg, Wendling), others stress the erosion of worker autonomy by separating control and execution over the labour process (Braverman, Panzieri, Pasquinelli), and yet others highlight the tendencies of modern technologies to wear human bodies and minds out beyond their natural limits (Rabinbach, Saito, Berardi). However, less attention has been granted to imagining better, communist technologies or AI more specifically. Most efforts have focused on the few paragraphs in the “Fragment on Machines” that promise a post-work liberation from labour through mass automation. However, Marx only briefly mentioned this option in one draft text and abandoned the option immediately afterwards. This choice should make us question Marx’s resolve for a “fully automated luxury communism” (Bastani) and look for alternative modes of worker-centric technological development. What institutions can reconcile new technologies like AI with worker desires for autonomy and wellbeing? What technologies would result from such a revolution in technological design? In this paper, I will conduct a historical-reconstructive reading of Marx’ notebooks between specifically the writing of the Grundrisse in 1857-58 and Capital: Volume I in 1867 to extract his positive proposals for worker-centric technology and its implications for AI development.

By Massimiliano Simons (Maastricht University)

Contemporary philosophical engagement with computing and AI is largely framed within the domains of AI ethics and the so-called "empirical turn" in the philosophy of technology. This dominant approach emphasizes case-based ethical analysis of emerging technologies, focusing on the implicit values embedded in technical artifacts and design practices.

However, this is not the only way to philosophically approach computers and AI. Nor has it always been the prevailing mode. The philosophical reflection on computing has a broader and richer history, which includes alternative frameworks that deserve renewed attention. This talk contributes to a larger project that seeks to historicize Dutch philosophy of technology, questioning the commonly held narrative that situates its origin in the 1990s with the rise of empirical and ethically oriented methodologies. Such a narrative obscures earlier traditions that were active in the Netherlands.

One significant but overlooked tradition is a distinctly Hegelian strand of thought, which interprets the history of technology as intertwined with the development of rationality and human self-understanding. Central to this tradition is the provocative claim that the invention of the computer and AI represents a new epoch in human development—an embodiment of what Hegel called the Absolute Spirit.

This idea was first articulated by Dutch philosopher Jan Hollak (1915–2003), a scholar of Hegel, who began exploring themes of technology and cybernetics in the 1960s (see Coolen 1997). In several publications (1963, 1966, 1968), Hollak argued that the modern computer should be understood not merely as a tool, but as a manifestation of a new stage in the objectification of reason. His work inspired a group of Dutch philosophers, including Theo de Boer (1932–2021), Louk Fleischhacker (1936–2006), and Maarten Coolen (b. 1943), who extended this Hegelian framework in various directions. Fleischhacker, for example, applied it to early debates about labor and artificial intelligence (1987, 1989), while Coolen developed a philosophical anthropology centered on the idea that our understanding of computers reflects—and reveals gaps in—our understanding of ourselves (1987, 1992).

The influence of this tradition persisted into the later development of Dutch philosophy of technology. Notably, Peter-Paul Verbeek (2000), a central figure in the empirical turn, explicitly references Coolen as a representative of the earlier, "transcendental" tradition from which the empirical approach sought to differentiate itself (see also De Boer 1984).

The aim of this talk is to recover and recontextualize this largely forgotten but influential tradition of Dutch philosophical reflection on computing. By highlighting its anthropological and metaphysical dimensions, I argue that philosophy of AI need not be confined to the (ethical) analysis of technical artifacts. Instead, it can also serve as a rich site for philosophical anthropology—wherein thinking about AI becomes a means of thinking more deeply about ourselves.

By Adam Takács (Budapest Metropolitan University)

This paper advances a deliberately non-standard claim: the philosophical significance of artificial intelligence does not primarily lie in cognition, agency, or ethics, but in a transformation of the historical conditions of knowledge itself. Drawing on Michel Foucault’s early archaeological project—especially The Order of Things—I argue that AI should be understood as an emergent episteme: a reconfiguration of the rules that govern what counts as knowledge, how it is produced, and how it circulates.

Foucault’s notion of savoir designates not individual acts of knowing, but historically specific formations that delimit the field of the thinkable. These formations are neither reducible to technologies nor to scientific theories; rather, they define the implicit conditions under which objects, subjects, and truths can appear. On this account, the transition between epistemes is not a gradual evolution but a discontinuous shift in the organization of knowledge. It is within this framework that Foucault famously diagnosed the possible “disappearance of man” as a figure of knowledge in modernity.

This paper proposes that contemporary AI signals such a discontinuity. Instead of treating AI as a tool that enhances human cognition or as an object of ethical evaluation, I suggest that it reorganizes the very processes of knowledge production, validation, and application. Machine learning systems, generative models, and large-scale data infrastructures do not merely extend existing epistemic practices; they redistribute epistemic authority, decenter human interpretive agency, and introduce operational criteria of truth grounded in statistical correlation, optimization, and prediction.

This shift produces a fundamental tension within the philosophy of technology. Prevailing frameworks—whether phenomenological accounts of human–technology relations or critical theories of automation—continue to presuppose a human-centered episteme in which technology mediates human knowledge. By contrast, the Foucauldian perspective developed here suggests that AI marks a transition toward an epistemic configuration in which knowledge is increasingly produced within non-human, infrastructural systems that operate below the threshold of experience and intention.

By reframing AI as a “historical episteme” rather than as a problem of cognition or that of a technological innovation, the paper contributes to ongoing debates about the status of AI—tool, agent, or infrastructure—by displacing the question altogether. What is at stake is not what AI is, but the historical reorganization of what can count as knowledge in its contemporary setting.