Learning from Slime Mold

This is an amazing documentary about the research computer scientists are doing on slime mold. They are using the slime mold to solve maze and networking problems and even operate robots.

In one experiment, oat flakes were placed on a dish to represent the major cities around Tokyo and the slime mold was placed in the corresponding location for Tokyo. The mold created a network between the oat flakes that was strikingly similar to Tokyo’s rail system. My complaint about the experiment was that it did not seem to take topography into consideration, but I looked into it and they used light to simulate mountains, water, and other obstacles. Neat!


Alan Turing

Alan Turing
Alan Turing

Alan Turing was an English mathematician who is regarded as the father of computer science and artificial intelligence. He is renowned in the computer science community, but he remains unknown by much of the general public. Here are just a few interesting facts about this truly remarkable man:

  • In his early 20’s, he developed the Turing machine, an imaginary device consisting of an infinitely-long tape divided into cells, a way to read and write symbols on the tape and move the tape left and right, a finite set of instructions, and some sort of memory to store the state of the machine. With this simple machine, one can simulate the logic of any mathematical concept or computer algorithm.
  • He built a machine that cracked German ciphers during World War II.
  • After the war, he was arrested for homosexuality (illegal in Britain at the time) and underwent chemical castration.
  • He came up with a way to determine whether or not a machine is intelligent, known as the Turing test. In a Turing test, a human converses with machines and other humans (unaware of which he/she is conversing with). If the human is unable to tell a machine from a human, the machine is believed to be intelligent. There is an annual competition known as the Loebner Prize that uses the Turing test to award chatterbots for human-like behavior.

If you are interested in learning more about Alan Turing, Radiolab did a podcast on him called “The Turing Problem” yesterday.


A few days ago, I saw the following video in which two chatbots talk to each other. I think my favorite part of the video is when one bot says, “Together we are robots” and the other bot responds, “I am not a robot. I am a unicorn.” Heehee. 😛

The video inspired me to find some other chatbots to talk with each other. Immediately, I thought of SmarterChild, a bot on AIM (AOL Instant Messenger), but SmarterChild was taken down in May 2009. So, I searched online and found Cleverbot. Perfect! Some funny snippets from the bots’ conversations are below:

Continue reading Cleverbot

Jeopardy: The IBM Challenge

For the first time ever on Jeopardy, a machine (named Watson) is competing against human opponents… and they aren’t just any humans, either. Ken Jennings holds the record for the most consecutive Jeopardy wins. Watson’s other opponent, Brad Rutter, has won more money than anyone else in Jeopardy’s history (over $3.2 million dollars).

Jeopardy’s IBM Challenge started last night and will continue tonight and tomorrow night. My husband and I eagerly watched last night’s episode to see how Watson would fare. He did a decent job and ended the night tied with Brad Rutter at $5000 apiece. Ken Jennings had $2000. Watson, however, made a humorous mistake last night when he buzzed in with “the 1920’s” after Ken Jennings had also just incorrectly answered ” the 20’s.” Whoops! 😛

What surprised me was that Watson is not connected to the Internet, so I wonder what Watson’s information database is like. I think it would be interesting to see what he can do with access to the Internet. Also, Watson is currently given the clues in text form. I don’t know if Watson gets the clues immediately, only after Trebek has finished reading them, or if there is some other sort of delay before Watson gets the clue. If Watson gets them immediately, it obviously gives the machine an advantage, because it can begin searching for an answer before the human contestants understand the clue. Human players, on the other hand, would have an advantage over the machine if they are able to see and start thinking about the clues before the machine gets them. It seems like a good solution would be to make Watson able to hear and interpret the clues audibly as the clues are being read and then be given the clues in text form as the clue is being read or after the clue has finished being read. On the other hand, I am sure that any good Jeopardy player must speed read through the clue, rather than listen to Trebek read it. *Shrugs*

Anyway, it was definitely fun to watch last night’s episode and I look forward to watching the two remaining episodes of man vs. machine. 🙂 As someone mentioned on Twitter last night, though, “IBM missed a HUGE comedic opportunity by not programming Watson to sound like Sean Connery.” 😛

EDIT: I found a PBS NOVA program on Watson and I would recommend watching it if you have an hour of free time (I :heart: PBS). Apparently, Watson’s information database consists of 10 million documents (mostly downloaded from the Internet), including encyclopedias (Wikipedia and others), dictionaries, thesauruses, IMDB, The New York Times, The Bible, etc. Watson does not have access to his opponents’ incorrect answers, which is why he repeated Ken Jennings’s response (he does, however, get to hear the correct answers). And, to answer my question from above, Watson gets the clues as soon as they show up on the board.

EDIT 2: Haha, someone suggested a final Jeopardy question of: “What word is displayed in the following captcha?” Funny stuff.