ARENA 7.0 Impact Report
The impact report from ARENA’s previous iteration, ARENA 6.0, is available here.Summary:ARENA 7.0 took place…
Category: Opinion
The Federal AI Policy Framework: An Improvement, But My Offer Is (Still Almost) Nothing
The Federal AI Policy Framework has been released. Well, it is a four page outline.…
Confusion around the term reward hacking
Summary: "Reward hacking" commonly refers to two different phenomena: misspecified-reward exploitation, where RL reinforces undesired…
The Distaff Texts
Though I spend most of my time studying what is labelled “history” in some manuscripts…
Untrusted Monitoring is Default; Trusted Monitoring is not
These views are my own and not necessarily representative of those of any colleagues with…
Why I am not buying IPv4 addresses as an investment
2026-01-17 Disclaimer Quick Note I did the actual research back in 2024-10. I polished the…
Positive-sum interactions between players with linear utility in resources
Sometimes people say things like "If the humans and AIs have linear utility in resources,…
No, we haven’t uploaded a fly yet
In the last two weeks, social media was set abuzz by claims that scientists had…
The Case for Low-Competence ASI Failure Scenarios
I think the community underinvests in the exploration of extremely-low-competence AGI/ASI failure modes and explain…
A List of Research Directions in Character Training
Thanks to Rohan Subramani, Ariana Azarbal, and Shubhorup Biswas for proposing some of the ideas…