Matching it to Minsky's model really shows the limitations of current AI - but I'm going to read more about the validity of Minsky's model before setting any thoughts in stone.
Thanks for sharing that.
It's definitely worth reading more about Minksy's models. I have a lot of references in-line, but here are a few more starting points. The near-final-draft version his book Emotion Machines is available for free on his MIT site[0] and WashPo gave it a very positive review[1].
It would be interesting to investigate some criticisms of the model, though I haven't found any that specifically target the 6 layer model.
[0]: http://web.media.mit.edu/~minsky/Introduction.html [1]: http://www.washingtonpost.com/wp-dyn/content/article/2006/12...