Wednesday, October 29, 2014

Moravec's Paradox

This isn't exactly math, but it's artificial intelligence (AI), and that's close enough... especially since a few posts back I wrote about IBM's "Deep Blue" and its 1997 defeat of chess grandmaster Gary Kasparov (at the time, a long-held goal of AI). Well, Moravec's paradox is the interesting idea that advanced or high-level reasoning and logic is much more easily mimicked by a computer system than are low-level sensori-motor skills that have evolved over millions of years... it's easier for a computer to learn to play chess, than to recognize human faces. This is one of those things that is fairly obvious when you stop to think about it... but, we often don't stop to think about it!
Here's what Steven Pinker wrote in "The Language Instinct":
“The main lesson of thirty-five years of AI research is that the hard problems are easy and the easy problems are hard. The mental abilities of a four-year-old that we take for granted – recognizing a face, lifting a pencil, walking across a room, answering a question – in fact solve some of the hardest engineering problems ever conceived…. As the new generation of intelligent devices appears, it will be the stock analysts and petrochemical engineers and parole board members who are in danger of being replaced by machines. The gardeners, receptionists, and cooks are secure in their jobs for decades to come.”   
A more recent blog piece applies the paradox to Google's self-driving cars, a creation I've certainly had trouble comprehending, given the countless issues/variables involved:

[p.s. -- actually, where are the dang flying jetpacks I grew up believing we would all have by now... forget the cars Google, I want my personal commuting jetpack!]

anyway, below, another somewhat provocative post applying Moravec's paradox to brain processing:

No comments: