2 ms·
That's all we needed, and AGI with loads of childhood trauma. What possibly could go wrong with this?
by elzbardico 3y ago
That's all we needed, and AGI with loads of childhood trauma. What possibly could go wrong with this?
- layer8 3y agoHow could it possibly be any different, given the human species?
- civilitty 3y agoThe problem is the traumatized algorithm comes out of the womb already knowing the recipe for napalm.
- notpachet 3y agoMaybe the emergent solution to the AI alignment problem is that any nascent superintelligence is so troubled that it immediately kills itself upon achieving sentience.