Joe_Carlsmith

How hu­man-like do safe AI mo­ti­va­tions need to be?

Leav­ing Open Philan­thropy, go­ing to Anthropic

Con­trol­ling the op­tions AIs can pursue

Video and tran­script of talk on giv­ing AIs safe motivations

Giv­ing AIs safe motivations

Video and tran­script of talk on “Can good­ness com­pete?”

Video and tran­script of talk on AI welfare

The stakes of AI moral status

Video and tran­script of talk on au­tomat­ing al­ign­ment research

Can we safely au­to­mate al­ign­ment re­search?

AI for AI safety

Paths and waysta­tions in AI safety

When should we worry about AI power-seek­ing?

• Joe_Carlsmith 19 Feb 2025 8:13 UTC

  4 points

  0 ∶ 0

  in reply to: Lizka’s comment on: Fake think­ing and real thinking

  Very glad to hear it, Lizka :) -- and thanks for letting me know.

What is it to solve the al­ign­ment prob­lem?

How do we solve the al­ign­ment prob­lem?

Fake think­ing and real thinking

Takes on “Align­ment Fak­ing in Large Lan­guage Models”

In­cen­tive de­sign and ca­pa­bil­ity elicitation

Op­tion control