descriptive claim
An RL training step changes relatively few model parameters.
Dwarkesh Patel · The next big breakthrough will be AIs learning on the job · 26 Jun 2026
In their words
The transcript segments this claim was taken from
“That's why actually very few parameters are changed.”
“during an RL training step.”Play from 14:38Report a misattribution
95%how strongly it was put, not whether it is true
85%importance to the conversation
99%confidence the claim was read correctly