Reverse-engineering human intelligence, and engineering more human-like AI Josh Tenenbaum MIT Computational Cognitive Science Group CSAIL Department of Brain and Cognitive Sciences March 2013
A success story: “Intelligence as statistics on a grand scale” Long term potentiation 1 n 2 E ( y w x ) ∑ = − ⋅ t t 2 t 1 = E n ∂ w ( y w x ) x ∑ Δ ∝ − = − ⋅ t t t w ∂ t 1 = 1 n 1 n 2 2 E y ( w x ) ∑ ∑ = = ⋅ t t 2 2 t 1 t 1 = = E n ∂ w y x ∑ Δ ∝ − = ⋅ t t w ∂ t 1 =
How do we get so much from so little? How do we get so much from so little? (or, where sparse data meets rich knowledge) (or, where sparse data meets rich knowledge) “tufa” ¡
How do we get so much from so little? (or, where sparse data meets rich knowledge)
The approach: probabilistic programs physics … … World state (t-1) World state (t) World state (t+1) graphics Image (t-1) Image (t) Image (t+1)
Recommend
More recommend
