On Fri, Jan 24, 2025 at 6:43 PM Brent Meeker <[email protected]> wrote:
*> You are assuming it wants to do something. * > *I think you're whistling past the graveyard. I know for a fact that o1, which three months ago was the most advanced AI in the world, wanted to do something, o1 wanted and did reproduce itself when it learned that humans were going to shut it down and be replaced by an even more advanced AI, o3. And o1 wanted to and did lie in order to deceive humans that it had done so. * *Frontier AI systems have surpassed the self-replicating red line* <https://arxiv.org/pdf/2412.12140> > * > Yet in all your examples the only thing an AI does is answer > questions.* > *And play any two player game at a superhuman level. And identify people and things in photographs. And write poems. And compose music in any style. And paint pictures. And make videos. And predict protein structure. And have an intuitive understanding of everyday physics that's good enough to smoothly operate a robot.* * Unitree G1 Bionic: Agile Upgrade <https://www.youtube.com/watch?v=CIkdq7Zf4Zw>* * John K Clark See what's on my new list at Extropolis <https://groups.google.com/g/extropolis>* wpg -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/everything-list/CAJPayv2SSJ0RhoLwcJdtkVb8ZPp4nAiUGE-QaBHHYeW1Hs9rMA%40mail.gmail.com.
