@HaydnBelfield
Haydn Belfield
3 months
The US AISI would be extremely lucky to get Paul Christiano - he's a key figure in the field of AI evaluations & literally the inventor of RLHF. UK AISI is very lucky to have Dr Christiano on its Advisory Board
Tweet media one
@dkaushik96
Divyansh Kaushik
3 months
I’m going to add some extremely important context this article is missing. The EO specifically asks NIST (and AISI) to focus on certain tasks (CBRN risks etc). Paul Christiano is extremely qualified for those tasks—important context that should’ve been included here. Another…
8
10
139
2
9
128

Replies

@StephenLCasper
Cas (Stephen Casper)
3 months
@HaydnBelfield *nitpick, but Paul didn’t invent RLHF. For example, see this survey which came out shortly after the DRLHP paper for some history. For instance, TAMER is a form of RLHF, but it was being used almost 10 years earlier. What Paul et al. did was actually getting RLHF to be useful.
1
0
15
@HaydnBelfield
Haydn Belfield
3 months
@StephenLCasper thats interesting context, cheers
1
0
2
@moreisdifferent
Dan Elton
3 months
@HaydnBelfield Wow, the UK team is pretty based.. and well rounded.
0
0
1