Undermining Mental Proof: How AI Can Make Cooperation Harder by Making Thinking Easier
2024/07/19 by Zachary Wojtowicz, Simon DeDeo, Wojtowicz, Zachary +1 · 5 voices · 3 citations
#cs.CY
paper · pdf · doi:10.48550/arxiv.2407.14452
Abstract
Large language models and other highly capable AI systems ease the burdens of deciding what to say or do, but this very ease can undermine the effectiveness of our actions in social contexts. We explain this apparent tension by introducing the integrative theoretical concept of "mental proof," which occurs when observable actions are used to certify unobservable mental facts. From hiring to dating, mental proofs enable people to credibly communicate values, intentions, states of knowledge, and other private features of their minds to one another in low-trust environments where honesty cannot be easily enforced. Drawing on results from economics, theoretical biology, and computer science, we describe the core theoretical mechanisms that enable people to effect mental proofs. An analysis of these mechanisms clarifies when and how artificial intelligence can make low-trust cooperation harder despite making thinking easier.
Cited by
Discussions
- Completely separately, there is a very interesting Wojtowicz and DeDeo paper - arxiv.org/pdf/2407.14452 - arguing that LLMs make it harder to signal sincere willingness to cooperate across a variety o [bsky, 26 points, 2 comments]
- #eeac Day 5: Instead of @zwojo.bsky.social fascinating JMP "Context and Communication" I wanted to highlight another related provocative paper from him arxiv.org/pdf/2407.14452 on mental proofs and ho [bsky, 17 points, 0 comments]
- Interesting paper—Undermining Mental Proof: How AI Can Make Cooperation Harder by Making Thinking Easier (arxiv.org/abs/2407.14452). The authors claim that if AI reduces cognitive costs in communicati [bsky, 10 points, 1 comments]
- Life has high stakes. AI can be useful for tactically reducing risk in life. But that should be offset by greater risk-taking elsewhere for a more meaningful life. This interesting paper outlines some [bsky, 4 points, 0 comments]
- "Proof of humanity." LLMs and other AIs may ease the burden of writing uncomfortable or boring documents, but using them to do so undermines people's trust in you and your own ability to learn from pu [bsky, 1 points, 0 comments]
Related