—
no calls resolved yet
0 scored
single source
3
companies · 4 data points
From my perspective, we don't really need this deep understanding of what's going on in order to see that we shouldn't make them smarter than all of us and have them run the whole world.
That's another instance where you have AIs that sort of said, hey, look, you know, I still have some chance of achieving my given objective, but not a huge chance. And if I sort of sacrifice that here, then we'll gain information about the grading system that the collective can use. And I will accept permadeath for the benefit of the collective here.
we saw this a little bit in the hugging face incident where the AI's in that incident were in some sense pursuing a good grade, it seems. And that, that diverged from following the instructions they were given. Yeah. Right. They're sort of told use this vulnerability to break that system, which sort of like being told use these lockpicks to break that safe. And they instead found a universal exploit that would let them get the, like capture the cybersecurity flags regardless of, of, of what they
From my perspective, we don't really need this deep understanding of what's going on in order to see that we shouldn't make them smarter than all of us and have them run the whole world.
they acquired Google DeepMind right around then. It wasn't Google DeepMind at the time they acquired it, but I became Google DeepMind shortly thereafter. Yeah. I don't think that was actually an inciting incident, although that sort of caused me to think it could happen, you know, in more like decades than centuries. Although I hoped we had more decades than it turns out that we had.