People working in probabilistic programming can be seen as working towards AGI, in some ways. I'd take a look at the work done by Tenenbaum at MIT and Goodman

People working in probabilistic programming can be seen as working towards AGI, in some ways. I'd take a look at the work done by Tenenbaum at MIT and Goodman at Stanford. There are also lots of academics related to them in different places. You can reconstruct this graph by looking at Google Scholar citations and co-authorships. A great starting point is: http://probmods.org/ http://probmods.org/