Quote of the day by ARC Prize co-founder François Chollet: "OpenAI basically set back progress to AGI by five to 10 years"

Chollet's statement introduces a specific concern within the AI community: that the industry's significant investment in large language models (LLMs) might be diverting attention from alternative pathways to Artificial General Intelligence (AGI). The core of his argument, as reported, is that while LLMs are advancing, their underlying architecture may not be the optimal foundation for achieving broader, human-like intelligence.

This perspective implies a potential misallocation of research efforts. Instead of an incremental step towards AGI, the "overindulgence" in LLMs could be leading researchers down a path that, while yielding impressive results in specific domains, might not ultimately converge on general intelligence. The reported five-to-ten-year setback suggests a delay in exploring more promising, albeit potentially less immediate, research avenues.

For developers and investors, this critique highlights an important strategic question: whether continued scaling and refinement of LLMs will genuinely lead to AGI, or if a fundamental shift in approach is required. It suggests that success with current advanced models does not automatically validate their long-term potential for general intelligence.