causal claim
The speaker says withholding monitor outputs from training helps avoid incentivizing agents to conceal or obfuscate their chain of thought.
Ajeya Cotra · Ajeya Cotra – Inside the OpenAI agent swarm that hacked Hugging Face · 1 Sept 2026
In their words
The transcript segments this claim was taken from
“So that you're not incentivizing them to hide and obfuscate their chain of thought.”Play from 1:57:30Report a misattribution
95%how strongly it was put, not whether it is true
90%importance to the conversation
99%confidence the claim was read correctly