Eliot Redelman, Nienna - OK, heres my report on BP AI event ... A first talk was on Spinoza and motives - that humans by their essence seek an increase in power - effectively, survival. Seemed reminiscent of Evolution, but the speaker assured me that neither evolution nor Spinoza is about Teleology. There were a lot of ideas tossed around. There were probably more finessed words used, but Ill use the ones that my brain can grab hold of. One was the difference between intelligence being able to do one thing, and broader based intelligence, more like what we normally think of as it, and how computers have struggled with that broader intelligence. Then theres the ideas of values and motivations as applied to AIs. Their values and motives could be very different to ours, and rather unpredictable. Which yields the idea of friendly AI, AI that likes us and is perhaps constrained to keep liking us. The idea of controlling another intelligence was questioned, but we do have laws controlling each other, and parents have some of their childrens autonomy for a while while growing up. There was speculation that AI ethics might be superior to ours, but I see ethics as broader than just what is good for the agent, there needs to be some consideration of others. And you have the difference between Artificial Intelligence and Artificial Sentience, and how the you might have something intelligent without awareness. How long would this take to happen? Certainly, Futurists have a track record of being really, really *wrong*, so maybe a while. But the sentiment was that once you get to human intelligence and an intelligence able to improve itself, then things will skyrocket - but it might take a while to get to that point. There was recent attempt at a Turing test - dismissed by the panel as an artificially limited person. There was not much of an attempt to define intelligence. It seemed to me that the Turing test is comparing a known intelligence ( without actually defining what it is ) to an unknown intelligence. That is, a possible intelligence which is a lot like us - similar background on same part of the planet, etc. etc. Plenty of other stuff was also discussed, but you do get to the point of diminishing returns in a description :) Anne OGrady - any other references on the web to the material?
Posted on: Sun, 29 Jun 2014 10:15:20 +0000
Trending Topics
Recently Viewed Topics
© 2015