Call me crazy, but I’m much more scared of AI than I am of climate change 
-
Do you mean the kind of AI that exists already (bc it can be used to great effect by tyrants, for example) or do you mean “superintelligent AI”?
Artifical General Intelligence, which would level up to become Superintelligence very quickly.
-
-
I think the concept of Superintelligence is logically flawed. All the uniquely human attributes we call "intelligence" reduce to a single functionality: Creating new explanatory knowledge. AGI will have it, too. But we can understand its explanations, too. And it'll be fallible.
-
E.g.,
@elonmusk and@samharris can talk about "the control problem" while also worrying about the AI's "open-ended utility function". But isn't this a simple contradiction? An AGI's values will depend on its moral knowledge, which it will create as we do. It'll be a person. - 8 more replies
New conversation -
-
-
How likely do you think it is that the model of intelligence/minds/etc. required for this scenario to plausibly happen is correct?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
AGI isn't even a coherent concept as of now, much less a practical reality. Moreover intelligent isn't fearful. Motives are. AI of any form doesn't have motives.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Human understanding relies on inductive leaps. Computers are purely deductive machines. How do you get inductive leaps from deduction?
-
Induction doesn’t exist. It’s a myth. Knowledge is conjecture. See Popper, Conjectures and Refutations.
- 20 more 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.