Maybe for discussion over pizza:
Paralysis Of Analysis - http://www.npr.org/templates/story/story.php?storyId=13

the "Paralysis Of Analysis" was very interesting!

That was informative as always. Thank to Monica and Bert. For the future I would suggest to leave Q&A for the end of each presentation, as we obviously overloaded Bert with questions and didn't leave him enough room to follow his slides and present result of his research.

The discussion on what makes something funny was interesting from a philosophical view. As an AI enthusiast; I was hoping that Bert would show a working joke detector, and was willing to overlook the fact that jokes would need to be expressed in something other then normal English. In my opinion, a working demo, or examples of jokes and non-jokes that the system can identify, would give Bert's higher-level design credibility.

Apologies to Ms Atkins for my having provided a loud PA system and noisy crowd in the other room, during my talk, disrupting me and making it not fun for you. Also for rambling horribly and not knowing the literature of the field I've researched for 25 years. I agree with Mr. Rondeau that I should have instantly implemented a large complex software system to give him immediate gratification, and was at fault in not demoing full human NLP for his pleasure, since it is so simple to do.

Bert jests. This one is, I believe, "Just Triumph" :-P

Of course, it's as unnecessary for him to apologize as it is for me as facilitator. TechShop gracefully lets us use a meeting space that seats 70 with an HDTV projector twice a month. For free. Which means I as a facilitator can't ask for my money back. Neither can you in the audience. This was the second worst of four bad events out of 75.

Andrew should realize that we need to know what to do before doing.
Debate twice, code once.