CAPTCHA Killer
A few posts ago I mentioned that we had put in a CAPTCHA verification thingy on our article comments form. For those that are unfamiliar with CAPTCHA, the acronym is one that the folks at Carnegie Melon University decided was a good name for those verification images that have a string of letters and numbers (often distorted) that you must type into a form to prove you are human. In case you are wondering, the acronym stands for Completely Automated Public Turing Test to Tell Computers and Humans Apart. Those not utterly fascinated with large, geeky brains should know that Alan Turing was a computer scientist that did research into ways to tell the difference between humans and computers back in the 50's.
wanted to post about this because I thought it was interesting, and also a bit funny. Our publisher, Kerry, had come across an article in the New York Times about this topic, entitled A Dog or a Cat? New Tests to Fool Automated Spammers, which discussed CAPTCHA technology and the people working on both sides of the fence.
What I thought was interesting was the mention of a Russian kid who spends his time working with Optical Recognition Software (OCR) to crack CAPTCHA implementations. The article talks about how, in a rather neutral fashion, there are folks out there working to get automated systems to be able to recognize the visual message presented in an image, and then also correctly input it into a form. As noted, this can be beneficial to both spammers, who would love to get through all CAPTCHA barriers, and security experts, who constantly need to know where vulnerabilities are and also what to watch out for.
I suppose in my mind, it's also important for computer science. I'm curious about what Turing would say when the statistics of average human users that cannot pass a Turing test is so high. It would seem that advances in this kind of technology, which for now seems to have it's roots in spamming, could really be useful for other types of applications. Then again, it would also seem that the above referenced article, while speaking within the context of the all-annoying Internet spam subject, does shed some light on a question that will be more and more pressing as computer hardware capabilities improve, what is (or should be) the difference between a computer and a human?
It seems like a funny question out of a movie from when we were kids, but I will put this one out there for the ethicist. What happens when an affordable computer (that everyone can experiment with) has the same hardware capabilities as a human? While on the one hand that leads to a very strange conversation about some abstract things, I like to stick with the simple answers, drawing from what I know. So here is the answer that I have. When computers can do as many calculations as a human brain can, then we will be defenseless against spam. Yep, I'm less concerned about a computer being self-aware and more concerned about it's super-human ability to send spam. After all, if the average adult can't pass a CAPTCHA test one out of every four times today, we are going to have problems in the future.
