Sur l’évolution technique de l’intelligence : une histoire artificielle

Sur l’évolution technique de l’intelligence : une histoire artificielle

🎙 David Bates 👥 149K 📅 October 24, 2025 ⏱ 36 min 👁 3K 📄 expert opinion 🧭 2026-08-03
Available in: English (current) Français

Keywords

intelligenceartificialhistorytechnologyautomation

Summary

David Bates, professor at UC Berkeley, delivers a lecture at the Collège de France’s 2025 symposium on forms of intelligence. He proposes an ‘artificial history’ of natural intelligence, arguing that since the scientific revolution, the mind has been understood as intertwined with bodies, machines, and prosthetic technologies like writing. This intertwining has led to viewing human intelligence in terms of automatism. In the contemporary context of hyper-technologization and increasing automation, this poses a threat to human independence, yet also demands crucial political and social decisions. Bates draws on a tradition that sees intelligence as resistance to automatism, referencing Bernard Stiegler’s concept of ‘désautomatisation’. He traces this history from Descartes and the concept of the automaton, through 19th and 20th century conceptions of mind, brain, and machines, early computing devices, cybernetics, and post-war human-machine interaction theories. He concludes with a critique of artificial intelligence, asserting that intelligence arises from the confrontation of the mind with technical and machinic forms of determination. The lecture is dense and philosophical, aimed at an academic audience.

171 words

Critical Evaluation

The lecture offers a profound and historically grounded perspective on the relationship between intelligence and technology. Bates skillfully weaves together philosophical, historical, and technical threads, demonstrating a deep familiarity with the works of Descartes, Stiegler, and others. His argument that intelligence has always been ‘artificial’ in the sense of being constituted through technical means is compelling and challenges simplistic dichotomies between natural and artificial intelligence. The discussion of automation as both a threat and a site of potential resistance is particularly timely. However, the lecture’s density and the poor quality of the transcription make it difficult to fully appreciate the nuances of his argument. The lack of clear structure and the rapid succession of references may overwhelm viewers unfamiliar with the intellectual history. Nevertheless, the content is rigorous and well-supported by references to key thinkers and concepts. The title accurately reflects the content, and the lecture delivers on its promise. The main weakness is the presentation, which could benefit from clearer organization and visual aids. Overall, this is a valuable contribution to the discourse on AI and human cognition.

179 words

Title / Content Match

The title accurately reflects the content, which explores the historical and technical evolution of intelligence, framing it as an 'artificial history'.

Quality & Reliability

8/10

The lecture is given by a professor at UC Berkeley, a recognized scholar in intellectual history. The content is based on historical and philosophical analysis, referencing key thinkers and concepts. However, the transcription is heavily garbled, making it difficult to verify specific claims and sources. The video is from a prestigious institution (Collège de France), which adds credibility.

Key Moments

Cited Sources

Concurring Sources

  • Collège de France - Colloque de rentrée 2025 — The lecture is part of this symposium, which likely includes related talks.

Contribution & Novelties

The lecture provides a novel historical framework for understanding intelligence as inherently artificial, challenging the nature-artifice dichotomy. It synthesizes insights from philosophy, history, and technology studies to argue that intelligence is constituted through technical means, and that this perspective offers a way to resist the dangers of automation.

Pour aller plus loin :

91 words

Radar Profile

The radar profile shows high scores in quality of information and reliability, reflecting the academic rigor and institutional backing. The quantity of information is moderate, and the technical level is high, indicating a specialized audience. The overall score is strong, with a slight dip in quantity due to the lecture's concise duration.

Reliability 8/10