On AI Risk
“I think sandboxing works to a point. Then I think monitoring works to a further point. Where we can look at what the models are doing, not only in their chain of thought, but we can do activation monitoring and try to really understand: are the models... Are we happy that the models are not deceiving us? And then, eventually, I think you need to believe with huge confidence that the models are truly aligned.”

Sam Altman · Undated

Title and organization are as of the date said.

Primary source verifiedCommenting outside a listed field of expertiseLoss of control

Source

Cite this quote

Altman, S. (n.d.). Cheeky Pint episode 39 transcript [Podcast episode]. Cheeky Pint. https://cheekypint.transistor.fm/39/transcript

More from Sam Altman

All 3 quotes by Sam Altman →

Others on the same topic

All quotes on loss of control →

Report an error in this quote