descriptive claim
The speaker says Meter’s alignment and training assessment considers pressures affecting chain of thought and evaluations, including whether agents are trained to deceive monitoring systems.
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
“talking about like,”
“you know, are there pressures on the chain of thought?”
“Like, are there pressures against the evaluations?”
“Are the agents being trained to fool the monitors?”
“That type of thing.”Play from 2:05:21Report a misattribution
90%how strongly it was put, not whether it is true
80%importance to the conversation
95%confidence the claim was read correctly