Here is my conversation with Francois Chollet and Mike Knoop on the $1 million ARCAGI Prize they're launching today.
I did a bunch of socratic grilling throughout, but Francois’s arguments about why LLMs won’t lead to AGI are very interesting and worth thinking through.
It was really fun discussing/debating the cruxes. Enjoy!
Check out ARCAGI Prize here: https://arcprize.org
Links:
Transcript: https://www.dwarkeshpatel.com/p/franc...
Apple Podcasts: https://podcasts.apple.com/us/podcast...
Spotify: https://open.spotify.com/episode/7bme...
Follow me on Twitter: https://x.com/dwarkesh_sp
Timestamps:
00:00:00 – The ARC benchmark
00:11:53 – Why LLMs struggle with ARC
00:19:43 – Skill vs intelligence
00:28:38 – Do we need “AGI” to automate most jobs?
00:49:11 – Future of AI progress: deep learning + program synthesis
01:01:23 – How Mike Knoop got nerdsniped by ARC
01:09:20 – Million $ ARC Prize
01:11:16 – Resisting benchmark saturation
01:18:51 – ARC scores on frontier vs open source models
01:27:02 – Possible solutions to ARC Prize