After OpenAI's blowup, it seems pretty clear that 'AI safety' isn't a real thinggizmodo.com295 pointsrntn3 years ago