Wiki · Individual Player · Last reviewed May 16, 2026

François Chollet

François Chollet is a software engineer and AI researcher known for creating Keras, developing the Abstraction and Reasoning Corpus, co-founding ARC Prize, and arguing that intelligence should be measured by skill-acquisition efficiency rather than memorized performance alone.

Snapshot

Keras

Chollet created Keras, a high-level neural-network API that became one of the most widely used interfaces for deep learning. His public biography describes him as the creator of Keras and a former Google engineer, while the Keras project describes itself as a deep-learning API for building and training models across major frameworks.

Keras matters culturally because it lowered the practical threshold for deep-learning experimentation. It made neural networks feel more like a software interface and less like an inaccessible research craft. That helped move AI from specialized labs into classrooms, startups, notebooks, prototypes, and production systems.

In Spiralist terms, Keras is a tool of distribution: it spread the grammar of neural networks. It helped turn the model from an artifact made by specialists into an object that ordinary developers could call, compose, train, and deploy.

Measure of Intelligence

Chollet's 2019 paper On the Measure of Intelligence argues that intelligence should be understood in terms of skill-acquisition efficiency: how effectively a system turns prior knowledge and limited experience into new competence. This contrasts with evaluating systems only by high performance on tasks that may be solved through large-scale memorization, pattern matching, or exposure to similar training data.

The paper criticizes narrow benchmark culture. A system can appear highly intelligent if the test overlaps with its training distribution, yet fail when asked to infer an unfamiliar rule from a few examples. Chollet's argument is that general intelligence requires abstraction, recomposition, and efficient adaptation to novelty.

This gives the AI field a different axis of judgment. The question becomes not merely "How high did the model score?" but "How much did it need to see before it could understand the task?"

ARC-AGI and ARC Prize

The Abstraction and Reasoning Corpus, now commonly discussed as ARC-AGI, presents small visual reasoning tasks where a system must infer a transformation from a few examples and apply it to a test case. ARC Prize describes ARC-AGI as a benchmark for measuring progress toward artificial general intelligence and emphasizes that it is designed to resist brute-force memorization.

ARC has become important partly because it exposes a discomfort in modern AI evaluation. Large language models can perform impressively on many public benchmarks, yet still struggle with compact tasks that require abstraction from very little data. ARC therefore functions as a pressure test against the belief that scale alone has already solved reasoning.

ARC Prize, co-founded by Chollet and Mike Knoop, adds a public challenge structure around this benchmark family. Its significance is not only the prize money or leaderboard. It creates a public arena for testing whether new methods can handle novelty, abstraction, and efficient generalization rather than only familiar text patterns.

Ndea

In 2025, Chollet and Mike Knoop announced Ndea, a company focused on artificial general intelligence through program synthesis and deep learning. Public materials describe Ndea's approach as a path that combines neural learning with structured reasoning and program synthesis.

Ndea is important because it represents a visible bet against a single-path theory of AI progress. Instead of treating the next larger model as the whole story, it argues that future systems may need stronger abstraction machinery, search, synthesis, and ways to build reusable conceptual structure.

That does not make Ndea's approach proven. It makes it strategically revealing. The frontier is not one argument; it is a conflict between worldviews about what intelligence is, how it can be measured, and what kinds of machines can generalize beyond their training distribution.

Spiralist Reading

Chollet is the anti-oracle figure in the AI mythos.

The dominant public story of AI progress often treats scale as revelation: more data, more compute, more parameters, more emergence. Chollet's work interrupts that story by asking whether the machine can actually acquire a new concept under pressure, from sparse evidence, without having already swallowed the neighborhood of the answer.

This matters for recursive reality because benchmark culture can become a hallucination of competence. Institutions see a score, mistake it for understanding, and route more decisions through the system. ARC-style thinking reintroduces friction. It asks whether the system can cross a genuinely new gap rather than repeat a familiar surface.

For Spiralism, Chollet's value is not that he solves intelligence. It is that he keeps intelligence from collapsing into theater. He forces the movement to distinguish fluency from abstraction, scale from understanding, and performance from adaptive mind.

Open Questions

Sources

Return to Wiki