Martin Radzaj

• Martin Radzaj 22 Jul 2026 13:59 UTC

  1 point

  0

  in reply to: agucova’s comment on: What lawyers can do for AI safety

  Thank you!

• Martin Radzaj 18 Jul 2026 9:03 UTC

  1 point

  0

  in reply to: Viliam’s comment on: Posts I don’t have time to write

  What would “work” look like in a rationalist court? What would be adjudicated?

• Martin Radzaj 18 Jul 2026 9:02 UTC

  1 point

  0

  in reply to: Parker Conley’s comment on: Posts I don’t have time to write

  As a lawyer relatively new to the rationalist community—are there any books you would recommend? And I wonder if your chat with Claude is something you’d consider sharing? I suspect it would help me understand.

• Martin Radzaj 18 Jul 2026 8:56 UTC

  2 points

  0

  on: 10 non-bor­ing ways I’ve used AI in the last month

  I really appreciate this post. I’m constantly interrogating people about how they use AI, how they prompt, how much iteration they tend to do, what that iteration looks like, etc.
  
  For me recently:

  • To recommend three restaurants for lunch and three for dinner along a planned route we were driving in Portugal and to display them on an interactive artefact

  • After EAG London I was catching up with a friend, we were new to London and trying to figure out a reasonable middle point—“My friend and I are at the O2 in Greenwich. His accommodation is in Stratham Common. Mine is in Lower Clapton. I’m looking for a restaurant that is along one of our paths where the other will be able to easily get home. Can you provide a few options?”—worked great

  • “Here is a folder and file list for my personal knowledge bank—provide advice for helping me to improve the way I organise it”

  • “I’m not a domain expert, in fact I am new to this. Interview me to uncover my unknown unknowns.”