While IBM's Watson obviously got a ton of attention for winning its big national TV challenge against two Jeopardy stars, apparently, IBM is taking the Jeopardy playing machine on tour, with a key stop being Congress. It got to play against a group of our elected officials... and one of them actually beat Watson. Say hello to Rep. Rush Holt -- who was a bit of a ringer, since he's actually a five-time Jeopardy champion. Holt apparently outscored Watson $8,600 to $6,200, though all the other Congressional Reps who went up against Watson were unable to master the computer. I guess this means that we should make Rep. Holt our new leader when the machines come to try to enslave us.

Today is the final game of Jeopardy! where the IBM supercomputer Watson plays against two of the best human players to ever compete on the show. Folks on the East Coast already know the outcome by now, so feel free to ruin the suspense in the comments below for those of us in later time zones. But whatever the outcome, Watson's performance has been pretty interesting to watch. And let's hope these supercomputers don't start playing thermonuclear war any time soon. In the meantime, here are some links on AI beating humans at other games and tests.

• Deep Blue won its first game of chess against Garry Kasparov in 1996. The computer didn't win the match that year, but it won the re-match in 1997. [url]
• Checkers was declared "solved" in 2007 by the Chinook project. Chinook was actually stronger than any human player by 1996, but it took a few more years for Chinook to realize checkers was a futile game (like tic-tac-toe) and retire. [url]
• A few years ago, the Polaris poker bot beat a few professionals at Texas hold'em. So be careful playing poker online... [url]
• The famous long bet between Mitchell Kapor and Ray Kurzweil has $20,000 riding on the question of whether or not AI will pass a Turing test by 2029. The bet started in 2002, and Kapor even suggested back then that a machine might win at a Jeopardy! game show. [url]
• To discover more interesting stuff on artificial intelligence, check out what the robots at StumbleUpon suggest. [url]  

By the way, StumbleUpon can recommend some good Techdirt articles, too.

Last month, we wrote that the IBM computing project Watson was ready to take on real Jeopardy contestants in February of this year (just a few years after the project first came to life). While the big test isn't until Valentine's Day, apparently they had a dry run, and things are looking pretty good for Watson, who beat Jeopardy champions Ken Jennings and Brad Rutter. It wasn't a complete domination. Apparently Watson just narrowly edged out Jennings, though Rutter couldn't keep up. Of course, between now and the real test, Watson can be tweaked. Jennings' and Rutters' brains are pretty much set, and I'd imagine that their ability to cram more useless trivia in their brains is outmatched by Watson.

The world of supercomputing and artificial intelligence has always been fascinated with games like chess, Go and poker -- where evaluating strategies for winning could be approached using vast computational resources. Another historical challenge for AI projects is the classic Turing Test, which requires an understanding of human communication in order to pass the test. So it's interesting to see IBM starting a project called Watson (named after IBM's founder, not Sherlock's assistant) to create a computer to compete on Jeopardy! against human contestants -- because doing so combines the goals of natural language processing and strategic game planning.

Obviously, though, Watson is more of a product demonstration than a serious research project to advance the field of computer science, but it's still a brilliant move, nonetheless, since the effort can be understood by anyone who watches popular game shows, and at the same time, the project studies a practical problem for algorithms. However, the game of Jeopardy! is somewhat trivial (no pun intended, honest) since Jeopardy's given answers all have matching questions -- and it doesn't really require true comprehension to guess (apparently, human contestants respond correctly about 85% of the time). Granted, Jeopardy! is a bit harder to play than Who Wants To Be A Millionaire, but Alex Trebek doesn't seem quite as computationally demanding as search engines can be (unless you count the audio/video Daily Doubles?). Still, IBM should be applauded for supporting its research as eye-catching PR campaigns -- while other companies like Microsoft are getting more attention for downsizing their research divisions.

Slashdot points us to an interview with Luis von Ahn (who we're a big fan of), where he talks about how spammers who are frustrated by various types of CAPTCHA tests have set up their own sort of "innovation prize," offering up somewhere in the range of $500,000 for software that can automatically pass CAPTCHA and reCAPTCHA reading tests (the things where you have to fill in a series of letters to sign up for a service or post a comment). As von Ahn points out: "If [the spammers] are really able to write a programme to read distorted text, great -- they have solved an AI problem." It is, effectively, an "X Prize" for optical character recognition. Not that we like to encourage spammers, but it is rather fascinating how the underground business seems to mirror the above ground innovation world as well.

Score one for humanity. After four grueling sessions of human vs. computer poker, the human players managed to outlast and outwit the bot. It was by no means an easy task, as the computer performed strongly on day one. But, by the second day, the human players seemed to have a pretty good feel for the way the machine played the game, giving them the leg up. Assuming that computer, dubbed Polaris, represents the vanguard of AI poker, it would seem that it'll still going to be a long time before computers can compete at the top level, as they can do in chess. This particular match offered the computer optimal conditions, as it faced an opponent heads up in limit poker; in a no-limit game with a full table of opponents (conditions that would make it even harder to calculate proper strategy), its performance would have likely been significantly worse.