The Reverse Turing Test

Not surprisingly, I spend a lot of time maintaining the Hati and Skoll website.  All websites nowadays have several levels of protection against “them evil spammers and hackers.” Recently WordPress, which is the fundamental engine beneath all of this added a simple “CAPTCHA” to the administrative login and this has worked wonders. A CAPTCHA, as I’m sure many of you know, is a simple question that, hopefully, only a human can answer.  In WordPress’ case its a simple math question like “1 + 3 = ?.” If you try to comment on Hati and Skoll you’re asked to read a little bit of graphic text. That’s another form of CAPTCHA.

What struck me was that the CAPTCHA is a “Reverse Turing Test.”  Alan Turing was interested in the concept/question of machine intelligence. He introduced his test in his landmark 1950 paper “Computing Machinery and Intelligence,  Turing begin the paper with the words: “I propose to consider the question, ‘Can machines think?'” He then morphs this question to one that is, perhaps, more accessible “Can machines do what we (as thinking entities) can do?”  We imagine an interaction between a human and a computer where the computer tries to prove to the human that he/she too is human by answering questions. As you can imagine there is a rich Science Fiction literature around this concept and for 65 years we have been intrigued with the question (and creeped out by the movies) how a machine could trip us up and prove to us that it is also human.

Well, fast forward to the modern day.  The question has essentially reversed. Indeed, the acronym CAPTCHA stands for “Completely Automated Public Turing test to tell Computers and Humans Apart.” Computers are not proving to us that they are human. Rather we are proving to them that we are human, remain human. The deeper  question is, of course, who’s in charge?