chanson

A couple days ago, Brad DeLong mentioned he wasn't quite sure what the Singularity was supposed to be, or how we'll solve the software problems involved in getting there.

I mentioned Cyc in his comments, and an anonymous commenter posted a report on a visit to Cyc-West by Vaughan Pratt of Stanford. It's an interesting look at the state of the project, which I consider to be one of the most promising of the ambitious AI projects started in the 1980s.

Of course, Pratt's report was written over a decade ago. I hear Cyc has evolved quite a bit since then, a time when an 8MB Symbolics Lisp Machine was studly. I'd like to see the results of asking a current Cyc the questions on Pratt's quiz.