It's alive! Security and Turing Tests.
One of the more interesting issues that have surfacing out of the murky security waters of the Internet is the newly-found relevance of what computer scientists call the Turing Test. This thought experiment, conceived of by noted British mathematician Alan Turing, seeks to determine the moment of sentience of an artificial being. It goes like this: you're in a room communicating by teletype (put new communications technology here). You have a lively conversation with someone somewhere else, using the equipment. At the end of a set period of time you can't tell whether they're a human being or an automaton. They've passed the Turing Test.
As far as I know, nothing has yet passed this quiz. A well-known 40-year old program called Eliza and its desendants often fool people for a few seconds. The Turing test that we might run into the most frequently is the "Completely Automated Public Turing test to tell Computers and Humans Apart" or CAPTCHA. This is the word in a picture that you have to type in to get access to some sites. The belief is that software cannot spot the word in the picture (not completely true) and therefore the reader is human. In fact, spammers are beating this system through the use of cheap or free human labor, coopting others to look at the picture and type the word. I think using the human being as a visual sensor does not break the spirit of the Turing.
I think that Turing tests are going to be big business and so intrinsic to our daily online life that we will shortly become very tired of them. Ultimately they are the only protection against spam. One way to look at the spam/counter-spam wars is that spambots and their enemies are quickly evolving as AI programs. They are getting quite adept at doing their job and may some day be viewed as alive in some sense, since they seem to have the main attributes; they reproduce and they protect themselves.
I believe that the very near future will see a war of semi-intelligent probes and hacking tools, spambots and viruses aht will attempt to punch through the hard shell of our personal computers and suck out our soft data goodness within. This battle moves too fast for humans to get involved. As anyone who runs a blog knows, spam has even worked its way down to blog comments and even spurious log entries.
It's a good time to be studying artificial intelligence. Once a discredited study, I expect it to get much bigger and more interesting as a spam fighting tool.
Posted on February 05, 2007
