factual

Explanation search may match model search in difficulty

The key hope of ARC's project is that finding a logical-style explanation for why a neural net works is of similar difficulty to finding the net itself, because the same way neural nets manage to embed discrete logical reasoning into continuous searchable spaces, explanations of that reasoning can be embedded and searched over in parallel — contrary to the common intuition that finding the net is much easier than explaining it.

factualpending

Speaker

Paul Christiano

Evidence Quote

“to the extent that neural nets are able to learn the rigid logical stuff at all, they learn it in the same way... it's possible to embed this discrete reasoning in the space in a way that's not too inefficient”

Source

Paul Christiano — Preventing an AI takeover— Dwarkesh Patel Podcast
Created: 6/13/2026, 3:36:05 AM

Your notes

Loading notes...