« Microslosh Strikes Again | Main | AI vs Minsky »
February 27, 2003
AI vs Minsky
Every since taking some AI subjects in Cognitive Science at UNSW, I have wondered how some leading AI researchers get away with it. The continual promises of solutions (just around the corner) and the bucket loads of money that are poured into "pure AI" are literally unbelievable. Well today there is an interesting article on slashdot where the discussion points out that there have in fact been some spin-offs from AI. The problem is that once some aspect of AI is "solved" it tends not to be be seen as AI anymore. In effect we move the boundaries. This does not really absolve people like Minsky who are still promising the "real thing" any day now but it does suggest that AI has had some benefits (expert systems, character recognition, voice recognition) which are now sort of independent of AI. The classic quote was about who the definition of intelligence ("using tools") has changed ("making tools") and changed ("self recognition") as we learn more and more about our environment.
The trigger for the AI article was a story about the Loebner/Minsky prize in Salon.
Posted by Ozguru at February 27, 2003 12:00 PM
Comments
Posted by: Ikh at June 7, 2005 11:57 PM
