Venkatesh Rao@vgr"Learning to learn" as distinct from open-ended adaptation is vacuous. It is generalization of some learning style rather than a meta thing11:04 PM · May 4, 20153 Retweets2 Likes
nikete@nikete·May 4, 2015Replying to @vgr@vgr there is some related work to this from a bayesian hierarchical models pov http://psy.cmu.edu/~ckemp/papers/kempgt10_learningtolearncausalmodels.pdf…
otis.@nftotis·May 5, 2015Replying to @vgr@vgr how would u distinguish between the two. Wouldn't learning to learn be self-deception bcuz ur becoming more adapted as a byproduct