rif retweeted
If an AI developer was on track to lose control of its internal agents, would anyone outside the company know?
Today, we are sharing METR’s vision for how we would work with AI developers to openly assess and share evidence about this.
A good time was had working with @ReillyHaskins02 on this. My first @METR_Evals publication!
rif retweeted
We run lots of evals at METR. Sometimes, agents attempt harmful actions.
I built a monitor that blocks suspicious tool calls until a human reviews them.
Writing out a case for why it's effective surfaced hidden assumptions. I'd recommend it to anyone building monitors!
After 19 years at Google, I'm delighted to announce I'm joining METR as a Member of Technical Staff. I'm excited about METR, their work to date, and their mission. Society needs independent expert orgs that can deeply understand and communicate about AI's capabilities and risks.
Has anyone considered the therapeutic applications of deepfakes? "I always wanted my dad to tell me I was good enough."
What direction do Islamic fans of Japanese monster movies face when praying? Towards Mecca-Godzilla.
@rifasaurousrex: On pirouline.com, they seem to have phased out the phrase "artisan rolled."
Just watched "Two hours of trippy, disjointed scenes vaguely reminiscent of moments from the book 'Ender's Game.'" 1.5 stars?
Sometimes, my dreams have a moral. Tonight it was "Never trust a sentient, life-sized Amazon Kindle." They'll kill us all. Smugly.
Today I gave a presentation to my boss's boss's boss's boss. My boss said he knew I rocked because top boss patted boss's boss on the back.









