One of the best posts I’ve here on LW, congratulations. I think that the most important algorithms that the brain implements will probably be less complex than anticipated. Epigenesis and early ontogenetic adaptation are heavily depended on feedback from the environment and probably very general, even if the ‘evolution of learning’ and genetic complexity provides some of the domain specifications ab initio. Results considering bounded computation (computational resources and limited information) will probably show that the ULM viewpoint cluster is compatible with the existence of cognitive biases and heuristics in our cognition http://www.pnas.org/content/103/9/3198
Thanks! That’s an interesting link. At the very lowest level, neural competition through local inhibitory circuits is a central mechanism used throughout brains. Of course that’s not the same thing as conflicting agents, but perhaps there is a general them of competition between subsytems at higher levels.
One of the best posts I’ve here on LW, congratulations. I think that the most important algorithms that the brain implements will probably be less complex than anticipated. Epigenesis and early ontogenetic adaptation are heavily depended on feedback from the environment and probably very general, even if the ‘evolution of learning’ and genetic complexity provides some of the domain specifications ab initio. Results considering bounded computation (computational resources and limited information) will probably show that the ULM viewpoint cluster is compatible with the existence of cognitive biases and heuristics in our cognition http://www.pnas.org/content/103/9/3198
Thanks! That’s an interesting link. At the very lowest level, neural competition through local inhibitory circuits is a central mechanism used throughout brains. Of course that’s not the same thing as conflicting agents, but perhaps there is a general them of competition between subsytems at higher levels.
Yes. That paper has been cited by Stuart J. Russell’s “Rationality and Intelligence: A Brief Update” and in Valiant ’s second paper on evolvability.