Man vs Machine: Watson Aftermath

  If your were one of the twelve million viewers last month watching IBM's supercomputer Watson competing on Jeopardy! against champions Brad Rutter and Ken Jennings, you saw a rather dominating performance from Watson. Watson finished the first game at $35734 and completed the three day, two game championship with a total of $77147 with Ken Jennings finishing second at $24000 and Brad Rutter with $21600. IBM claimed the million dollar prize for the championship, donating the money to charities World Vision and World Community Grid. Jennings took home $300000, donating half to VillageReach, and Rutter claimed the $200000 prize, donating half to the Lancaster County Community Foundation.

Ken Jennings finishing off Final Jeopardy with a bit of humor

 
  Watson lead a commanding performance through most of the games. Watson would only buzz in when it had a confidence level above a certain threshold but its buzz was lightning fast. The only times Jennings or Rutter were able to chime in were either when Watson had low confidence in its answer or when Rutter or Jennings anticipated the buzz and beat Watson to the punch. Watson also played very logically, if a bit bizarrely to human strategies. It would hop around the board searching for the Daily Doubles and when it found them it would bet very precise amounts based off of the current state of the game, wagering $6435 on an earlier Daily Double and $17973 on the Final Jeopardy on the third day.
  Not all of Watson's performance was perfect. Watson couldn't "hear" what the other contestants buzzed in with, if Ken or Brad rang in first and answered incorrectly, Watson would sometimes buzz in and give the same answer. At the end of the first game, I'm sure much to the chagrin of its creators, Watson's answer of "What is Toronto?????" to the Final Jeopardy category "U.S. Cities" was a bit off the mark. The presence of the five question marks displayed a very uncertain answer from Watson. The IBM creators said they didn't give much weight of choosing answers to the category title because Jeopardy! writers would often use jokes or puns in the titles. Unfortunately for Watson's human competitors, it wagered only $947 of its $36681 current first day total, not giving an ounce of respite for the challengers. For those interested, here is the answer that stumped Watson:

"This U.S. city's largest airport is named for a famous World War II hero, its second largest for a famous World War II battle."

  Overall, I was incredibly impressed by IBM's showing on Jeopardy! Watson showed a depth and understand of the human language that could have a great number of implications. Being able to parse a question and delve through terabytes of information and return an answer in a few seconds is a critical step towards an era of intelligent, helpful computers. IBM wanted to show off what they could do, and show they did.

Algorithms in Action: Genetic Algorithm

  Artificial intelligence uses a great number of algorithms to perform various functions required by the AI. An algorithm is a structured set of logical instructions to reach some solution to a problem. One such algorithm is a genetic algorithm, a search heuristic that mimics heredity to explore various solutions based on the fitness, or correctness, of each element in a generation. Heuristics are experience based approaches that utilize previous knowledge for problem solving or learning, such as a rule of thumb or trail and error. BoxCar2D has been created to give a very good visual on how the genetic process works.
  Genetic algorithms follow a flow not unlike biological genetics to simulate evolution of a system. An initial population is created, generally randomly, and each element of that population is set through a fitness test. After these tests are run, a percentage of the population that is deemed most fit is selected and allowed to "breed." This select division of the population mix portions of their code together to generate a new population of children that will hopefully improve overall fitness of the new population. Along with the mixing of the genetic code the algorithm can introduce mutations to a percentage of the next generation, randomizing different segments of the code to introduce new traits and to prevent an evolutionary dead end.

BoxCar2D in action

  In the BoxCar2D simulation, the user can specify a few traits of the car and watch it evolve. The fitness testing of the car is based on how far the car can go and how fast. The best few are selected and pass on their traits to the next generation and the pattern continues. The user can affect the selected traits by up or down voting cars and changing the mutation rate, which can affect body shape, wheel size, or wheel placement. The graph that populates gives the user a look at how the generations do on average compared to the fitness threshold for it. Take it for a drive and see how it works.

Man vs. Machine: Computer to play on Jeopardy!

  Starting on February 14th, Ken Jennings, Brad Rutter, and Watson will be playing a three day tournament with a million dollars on the line. Ken Jennings and Brad Rutter are the two top players of Jeopardy! fame, Jennings having the longest winning streak on the show and Rutter being the largest money winner after winning the Ultimate Tournament of Champions. But who is Watson? Not Sherlock Holmes' Dr. Watson, no, Watson is a supercomputer created by IBM.
  Watson is an advanced artificial intelligence designed to answer questions in a very human, very natural context. Using a complex system of algorithms to process natural language, hold, search, and retrieve terabytes of information, Watson can determine the answer to almost any question in a few short seconds. Outfitted with a buzzer, Watson is set to challenge Jennings and Rutter to an intense mind vs machine match on Jeopardy!

Ken, Watson, and Brad during a practice match on January 13th, 2011.