2 ms·

IMHE, it's been part of widespread discussions in the AI research and AI safety communities since the 2000s.