Re:A universal measure of intelligence
by bhousel on Friday January 28, 2011 @11:49AM
(#2152496)
Article is really light on details.. But they mention that they are basing their intelligence metric on an organism's Kolmogorov complexity. It sounds to me like they are quantifying randomness, not intelligence.
[ Parent
| on HN ]
Re:A universal measure of intelligence
by abecedarius on Friday January 28, 2011 @01:05PM
(#2152802)
I believe they meant the Kolmogorov complexity of the environment constituting the problem posed to the agent -- not the complexity of the agent itself. I don't have access to the paper, though. (Edit: the paper confirms this -- thanks, rubidium.)
Actual paper can be found at one of the authors websites: http://www.csse.monash.edu.au/~dld/Publications/HernandezOra....
Generally applicable hint: google the paper title. There's often a free pdf posted somewhere, usually at an author's webpage.