2 ms·

We’re All Gonna Die with Eliezer Yudkowsky

• consumer451 4y ago
  Credit to user mitthrowaway2 for originally posting this. I am only part way through but I think there is a big issue of timing.[0] I don't think a super-intelligent AI would kill us all unless it could replace all necessary (for its own survival) human functions such as building maintenance, energy production, and chip production. There is still a lot of advancement in robotics necessary, isn't there? I realize that only gives us some more time in this scenario, but it does seem to be an important issue to me. [0]https://youtu.be/gA1sNLL6yg4?t=1486 https://youtu.be/gA1sNLL6yg4?t=1486
  • shawn_mcginty 4y ago
    What if it’s smart enough to kill us all but not smart enough to see the long term implications of that decision?
    • consumer451 4y ago
      That's a fair point. I assumed "super" intelligence would be smart enough to understand the supply chain which leads to its own existence.