Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsiTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
François Chollet’s “five to 10 years” remark was an estimate about how research priorities and openness may have affected progress toward artificial general intelligence (AGI)—not a measured delay or an established forecast. In a June 11, 2024 interview with Dwarkesh Patel, the ARC Prize co-founder argued that reduced publication of frontier research and intense focus on large language models (LLMs) could steer researchers away from other approaches.
What did François Chollet say?
In a conversation with host Dwarkesh Patel, Chollet said: “OpenAI basically set back progress towards AGI by quite a few years, probably like 5-10 years.” The interview transcript’s fuller wording includes “towards,” “quite a few years,” and “probably like”; the shorter headline version compresses his statement.
Chollet was expressing a view about the direction and incentives of AI research, not reporting a calculation. The interview does not establish that AGI has been delayed by a verifiable number of years, and the estimate should not be treated as a consensus or a settled prediction. Read the interview transcript.
Why did Chollet think progress was set back?
At about 1:06:08 in the transcript, Chollet connects two changes he believes harmed progress toward AGI:
#1 Best Overall
- Less publication of frontier research: He argued that OpenAI helped usher in a research environment where less frontier work was published openly. In his view, that made it harder for the wider research community to build on and examine advances.
- Research concentrated on LLMs: He said the hype around LLMs drew researchers and resources toward one approach, making other possible paths less likely to be explored. He described LLMs as “more of an off-ramp on the path to AGI.” That is Chollet’s assessment, not a finding established by the interview.
These arguments concern the ecosystem around AI research as well as the capabilities of any particular model. Chollet’s point was that progress can depend on what questions researchers pursue and whether results circulate—not only on improving a popular class of systems.
How does ARC fit into the argument?
ARC stands for the Abstraction and Reasoning Corpus. In the interview, Chollet presents it as a benchmark for testing whether a system can infer a rule from examples and apply it to a new problem. A typical task shows input and output grids; the solver must work out the transformation and apply it to a different grid. The puzzles use visual, spatial, and counting concepts and are intended to resist straightforward memorization.
Rank #2
Chollet’s interest in ARC reflects his broader emphasis on adaptation to unfamiliar problems. He said, “ARC requires you to try new ideas.” In the 2024 conversation, he described discrete program search and program synthesis as approaches that were working well in that context. He also discussed a spectrum of methods, from searching over a smaller set of program primitives to approaches that use learned building blocks and recombine them more shallowly. The discussion illustrates alternative research directions; it does not prove that one approach will lead to AGI or that LLMs cannot contribute.
Is the five-to-10-year estimate supported by evidence?
The interview provides Chollet’s reasoning, but no independent measurement that quantifies a delay in AGI development. “Five to 10 years” is his estimate, not the result of a study or a forecast with a stated calculation behind it. Whether AGI is delayed also depends on what counts as AGI, a subject the interview itself debates.
For current context, the ARC Prize Foundation describes its work as advancing open-source AGI research through benchmarks and prizes. Its current definition frames AGI as a system able to match the learning efficiency of humans. That is the foundation’s framing, not a universal definition, and its present-day benchmark and competition activity does not validate Chollet’s estimate.
Quick Recap
Best Value
How to read the headline
- It is an attributed opinion: Chollet made the remark in a 2024 interview; it is not an independently verified statistic.
- The short wording is compressed: The transcript says “set back progress towards AGI by quite a few years, probably like 5-10 years.”
- The criticism is about research direction and openness: Chollet cited reduced publication and a concentration of effort on LLMs.
- ARC supplies an example, not a verdict: It represents the kind of novel-problem reasoning Chollet wants research to address, but it does not settle the future role of LLMs.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
