More realistic is previewing to see if they should keep it for themselves.

Spooked by Anthropic’s paranoia-inducing hype, of course.

“AI safety” has nothing to do with preventing psychosis, or addressing pollution or rampant electricity use (real problems). States are still banned - by Trump - from addressing those.

What is it? Well, it’s nothing, basically. It’s a fear of something that doesn’t exist, doesn’t have a path to existing, so deciding whether something is “safe” is bound to be based on which big tech corpo lobbies the hardest.