Raemon comments on The inordinately slow spread of good AGI conversations in ML

[Raemon]

I think one thing is that they had a different idea of what AI risk entails.

[Evan R. Murphy]

What idea would that be, misuse risks rather than misalignment risks?