How “reverse game theory” could keep AI agents in check
Mechanism design offers a behavioral approach to AI alignment that does not rely on understanding the complex internal workings of frontier models.
Reported by MIT Technology Review ·
What happened
AI safety researchers and economists are exploring "reverse game theory," or mechanism design, to create incentives that make following rules the most rational choice for AI agents. Proposed strategies include rewarding models for predicting another model's output to prevent sandbagging, as well as assigning models alternating roles as problem "solvers" and "auditors" to catch cheating.
Who it’s about
- MIT
- Princeton
- Stephen Morris
- Elad Hazan
- Helen Qu
Source
MIT Technology Review
How “reverse game theory” could keep AI agents in check
More on safety
Get a brief picked for you
Sign in to follow topics and companies, save stories and search the archive.
Sign in