Posts

AI Safety Polska community meet-up in Kraków — Representation Engineering 2024-12-06T03:28:34.194Z
GPTs' ability to keep a secret is weirdly prompt-dependent 2023-07-22T12:21:26.175Z

Comments

Comment by Marcel Windys on The Paris AI Anti-Safety Summit · 2025-02-14T02:05:08.651Z · LW · GW

It seems that a Vibe Shift is possible - if Trump's voters would realize what OpenAI and others are actually trying to build, they would be furious.

Comment by Marcel Windys on Stuxnet, not Skynet: Humanity's disempowerment by AI · 2023-11-07T00:35:32.447Z · LW · GW

If agriculture and transportation are fully automated, can't AI just deny military access to gasoline and electricity and then start to build data centers on farmlands?