We don't understand how to align human intelligence either. Wouldn't it be easier to start with that problem since you can actually do observations and experiments, rather than trying to hole-in-one the development of new technology using purely a priori reasoning?
-
-
Your suggesting simplicity? Oscar Wilde had a comment on such.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
But, yes I doubt it will, soonish. For my level of IQ, anyjow
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
No, it's mostly useless. - Functioning humans are a tiny region of mindspace, so they tend to independently "fill in the gaps" of what you asked them to do. Even baby AIs don't: https://docs.google.com/spreadsheets/d/e/2PACX-1vRPiprOaC3HsCf5Tuum8bRfzYUiKLRqJmbOoC-32JorNdfyTiRRsR7Ea5eWtvsWzuxo8bjOxCG84dAg/pubhtml … - Some classes of misalignment don't show up at current optimization power.
-
What's the evidence for radically different mind architectures besides conceivability? I'm aware that computer programs are deterministic based on exactly what's written. I still don't understand where "minds" enter into it. Not saying they don't, just show me the step.
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.