Moravec's Paradox

#1

Post by SongofEarth » Sat, Jun 10, 2023, 2:09 PM UTC

Re: Moravec's Paradox

#2

Post by alexmolas » Sat, Jun 10, 2023, 2:48 PM UTC

> Moravec wrote in 1988, "it is comparatively easy to make computers exhibit adult level performance on intelligence tests"

I wonder what made he think this in 1988. I guess he's referring to chess engines, but until a couple of years ago text generation was certainly not that good.

Re: Moravec's Paradox

#3

Post by haltingproblem » Sat, Jun 10, 2023, 2:56 PM UTC

Very interesting that Rodney Brook is linked to this in the Wikipedia article. Brook's subsumption architecture seems a logical evolution of this thought. Defer intelligence to the edges - the limbs - via subsumption behaviors. Also Brooks's famous paper "Elephants do not play Chess" [1] comes to mind. Fascinating to the see this various threads linked together. What a joy it would be to drop into conversations between Moravec, Minsky and Brooks!!

[1] https://people.csail.mit.edu/brooks/papers/elephants.pdf

Re: Moravec's Paradox

#4

Post by meltyness » Sat, Jun 10, 2023, 2:58 PM UTC

... I mean sure, if you require digital precision and reliability - had these guys not heard of a comparator?

Re: Moravec's Paradox

#5

Post by thunderbong » Sat, Jun 10, 2023, 3:21 PM UTC

I guess I'll get down-voted for this, but whenever someone asked me why I hate jogging, I replied, "I'm a software developer because I'm lazy!"

I enjoy weight and resistance training though!