Zach Stein-Perlman(Zachary Stein-Perlman)

Staged release

Deep­Mind: Eval­u­at­ing Fron­tier Models for Danger­ous Capabilities

OpenAI: Pre­pared­ness framework

An­thropic, Google, Microsoft & OpenAI an­nounce Ex­ec­u­tive Direc­tor of the Fron­tier Model Fo­rum & over $10 mil­lion for a new AI Safety Fund

OpenAI-Microsoft partnership

[Question] Cur­rent AI safety tech­niques?

ARC Evals: Re­spon­si­ble Scal­ing Policies

How to think about slow­ing AI

Cruxes for overhang

Cruxes on US lead for some do­mes­tic AI regulation

[Question] Which paths to pow­er­ful AI should be boosted?

[Question] Which pos­si­ble AI sys­tems are rel­a­tively safe?

[Question] AI labs’ re­quests for input

[Question] Boxing

Fron­tier Model Forum

My fa­vorite AI gov­er­nance re­search this year so far

In­ci­dent re­port­ing for AI safety

Fron­tier AI Regulation

AI labs’ state­ments on governance

Deep­Mind: Model eval­u­a­tion for ex­treme risks