19K views · 431 reactions | Dave Downing of CI Surfboards gives us a...

Global AI Insights on Instagram: "UC Berkeley Professor WARNS: AI Knows Self-Preservation 😱 AI safety researcher Professor Stuart Russell O.B.E.uses the King Midas myth to explain a core danger of modern AI: we’re often terrible at stating what we really want, and machines are brutally good at optimizing whatever objective we give them. Russell argues that classic AI worked when goals were simple (win at chess, win at Go). But in real life, “a good future” can’t be written down cleanly. That mi

umc.cmc.213sjmc7onm0nycpoilp9bjsu