P(house burning) ~ 0.24% - i buy home insurance.
I’m very optimistic about AI safety, I’d say P(doom) < 0.5% , that still requires insurance.
What are you doing?
Build towards a future where AI plan subsidies don’t exist.
How can you extract maximum efficacy from smaller, affordable models? Been putting a lot of thought (and work) into this over the past year.
Excited to share the fruits of that labor before the rug is pulled.