The day before the asteroid hits, people will still be writing thinkpieces called "But What About The Metaphorical Asteroid...Of Capitalism? Or Did I Just Blow Your Mind?"
-
-
What I'm really missing here is an argument or evidence that is it indeed feasible to cross the threshold between solving symbolic rule systems / concrete pattern matching tasks, and super(general)intelligence. AlphaGo isn't evidence towards that according to me.
-
Seems like a weird place to put the burden of proof. "Sure, cars went at 20 mph last year and 30 mph this year, but nobody can prove we can ever make a 40 mph car, so why worry?"
-
I agree it's possible that at some point the hitherto-consistently-improving wave of AI progress hits a mysterious brick wall and stops because our brains are magic and impossible to replicate, but I don't think it should be our default assumption.
-
Also, some of the stuff going on with machine vision now seems very different from just symbolic rules / concrete pattern matching. Ability to replicate vague artistic styles seems like good example of moving beyond the easily discretized. Obviously still far from human level.
-
What neural nets do is approximate a mathematical functions: a mapping from a set of bits to another. I can see the artistic styles used in the experiment being expressed as a mathematical transform: apply some filters, turn a few knobs on your image editor.
-
AI has not progressed constantly but rather by fits and starts. And each time the view of what was possible has been over-optimistic, on the "natural language solved by next summer" scale.
-
What we see currently, I think: the low-hanging fruits of a new technique (actually an old technique finally made practical) and the higher-hanging fruits of painstakingly combining it with other well established AI building blocks (Alpha Go).
-
Computing is, as a rule, not that surprising. Solving Go is certainly not outside the realm of the imaginable. Neural nets were conceived decades ago. But the vaguely-defined superintelligence, that is something else, more akin to making you 20 mph cars reach Mach 1.
- 8 more replies
New conversation -
-
-
This sounds random. My favourite part of ur piece is learning that it’s possible to circumvent Godel theorem if think in probability.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.