descriptive claim
The proposed reinforcement-learning process repeatedly runs AIs through virtual environments, training on successes and against failures.
Scott Alexander · AI 2027: month-by-month model of intelligence explosion — Scott Alexander & Daniel Kokotajlo · 3 Apr 2025
In their words
The transcript segments this claim was taken from
“But just running them through these environments again and again,”
“training on the successes,”
“training against the failures,”Play from 48:23Report a misattribution
85%how strongly it was put, not whether it is true
70%importance to the conversation
96%confidence the claim was read correctly