Work done in Josh Tenenbaum's lab. How to grow a mind: http://videolectures.net/aaai2012_tenenbaum_grow_mind/
Bongard Problems: somehow related to some core aspect of intelligence. It's some kind of maximization in some predicate space of the simplicity of a predicate subject to that predicate separating the two classes? Vague because I don't know :(
What is the internal representation scheme for representing spatial relationships and constraints in the world?
We learn through a curriculum, slowly. We don't get the entire world thrown at us all at once
It's so simple! :)
An awesome experiment from 1971 that shows that we are capable of mental rotation in our heads and that it works at constant rate as function of angle. In other words, we're running a tiny rotation procedure that modifies some kind of shape representation. More info on wiki: http://en.wikipedia.org/wiki/Mental_rotation
Why is there a white triangle here, when there isn't one? Somehow the explanation of the white occluding triangle being there is a higher probability configuration, because pacman style shapes and lines arranged exactly like that are too accidental.