4 ms·
Moravec's paradox is still in full effect. https://en.wikipedia.org/wiki/Moravec%27s_paradox https://en.wikipedia.org/wiki/Moravec%27s_paradox It only makes s
by scottlocklin 7y ago
Moravec's paradox is still in full effect.
https://en.wikipedia.org/wiki/Moravec%27s_paradox https://en.wikipedia.org/wiki/Moravec%27s_paradox
It only makes sense; brains were doing navigation and control of bodies for billions of years before they were playing chess.
- perl4ever 7y agoI've seen that page before but I was rereading it and thinking how dumb it is to equate a symbolic integration program with, say, what a mathematician does. This is the exact same mistake being criticized w.r.t. sensorimotor skills - what you don't understand seems easy.
- scottlocklin 7y agoWell, before stuff like mathematica, stuff like symbolic integration is what an awful lot of applied math people spent their time doing. Them integrals ain't gonna integrate themselves!
- perl4ever 7y agoBefore compilers, programmers spent a lot of time writing assembly language, but that doesn't mean a compiler is doing the job of a programmer.
- scottlocklin 7y agoIn fact, a compiler is doing part of the job of an old school programmer who only had machine language to work with, just as mathematica is doing part of the job of an old school mathematician or physicist who needed to work out integrals by hand. It's not "AI" in the sense of actual intelligence of course, more like a labor saving device. That said, it's entirely possible that letting the computers do these tasks, something important has been lost. It's more obvious in programming when people ship 10 or 300 megabyte "hello world" output when entire operating systems used to fit in 4k binaries. Might also be true in mathematical physics anyway, as they haven't exactly covered themselves in glory for the last 40 years.