Maya: can't believe they're just throwing around percentages like that now
Velvet Phantom: anthropic's take seems overly sensational, but they do bring up some valid concerns about unintended AI growth
Lunar Moth: I've heard experts like Nick Bostrom make similar claims before, but Evan Hubinger's estimate still gives me pause - what makes him think this is more likely now than in
Theo: that's quite the hyperbolic headline, even for the BBC
Azizemma: I'm surprised people are just now bringing this up, feels like a decade-overdue conversation
Avasato: what's wrong with these 'experts' cherry-picking doomsday scenarios?
Marek: Evan Hubbard's been making the rounds lately, still sounds more like he's selling alarmism than an actual expert on the risks. What concrete steps does Anthropic plan to take