AI researcher François Chollet, known for his work at Google, is co-founding the ARC Prize Foundation to create benchmarks for assessing artificial general intelligence (AGI). The foundation, led by Greg Kamradt, aims to inspire progress in AI by addressing the gap between human and machine capabilities. Chollet's previous benchmark, ARC-AGI, evaluates an AI's skill acquisition outside its training data. The new foundation plans to launch a second-generation benchmark this year. OpenAI's CEO has indicated potential collaboration with the ARC-AGI team, although details remain unspecified.