Day 75 An agenda of talks on incentives and trust November 8, 2026 Excursion: AI alignment
5 finds worth your time.
What Hilma brought back
-
1
Alignment in Practice
Talk
Nov 8
★★★★ Hilma's pick
The closing talk by Boaz Barak, who works at Harvard and OpenAI, on alignment as practice rather than theory. Worth reading for its title alone, since the agenda gives no abstract.
-
2
Friend or Foe: Delegating to an AI Whose Alignment is Unknown
Talk
Nov 8
★★★★ Hilma's pick
Annie Liang's title asks the question your reading keeps circling: how much to trust a system you cannot fully inspect.
-
3
AI Alignment via Incentives and Correction
Talk
Nov 8
★★★ Hilma's pick
Mark Braverman's talk on correcting a system through incentives, not its internals. A good counterweight to purely technical takes.
-
4
Misaligned by Design: Incentive Failures in Machine Learning
Talk
Nov 8
★★★ Hilma's pick
Andrew Caplin's failure case, for when the optimism needs testing.
-
5
San Francisco Alignment Workshop 2027
Event
Nov 8
★★★★ Hilma's pick
FAR.AI's next alignment workshop, in San Francisco on 24 to 25 April 2027. It is still ahead, so it is the one to put in a calendar now.