This New Scientist article gives an overview about how programmers are being goaded by X-prize-like rewards from spammers to crack CAPTCHAs - and, in the process, are providing a push towards fundamental breakthroughs in Artificial Intelligence.
The obvious and tongue-in-cheek conclusion drawn by the report is then "to start designing CAPTCHAs in a different way – pick problems that need solving and make them into targets to be solved by resourceful criminals."
The subtext of this cat-and-mouse game of advances in CAPTCHAs and the technology to crack them is actually a fairly simple task - to define something that uniquely identifies us as human. After all CAPTCHA stands for 'Completely Automated Public Turing Test To Tell Computers and Humans Apart' - and if someone is able to automate the solution then it no longer qualifies as a CAPTCHA.
In this escalating arms race, the demise of the CAPTCHA-as-we-know-it is probably very near. But the solution to stopping spam may not lie in moving the bar a bit further each time by repeatedly creating problems that need technological breakthroughs - or (as framed in the above article) problems that need an infusion of funds to motivate programmers to find a solution.
My bet is that the solution to a lasting CAPTCHA is not to focus on exceptional human capabilities (reading distorted text, correcting the orientation of pictures, etc.) but instead to focus on human frailties and chinks in the human mind that allow it to be fooled.
Visual and cognitive illusions - and magic - mask the perception of physical reality by exploiting our sensory and cognitive weaknesses - and could actually be the basis of next-generation CAPTCHAs aiming to establish how to identify humans from non-humans.
Of course, that would set off a race to simulate the mind's shortcomings and the creating of an artifical intelligence mimicking the neural basis of human intelligence. Again, my bet is that it wouldn't be as easy - or as near in the future - as cracking the human mind's specific capabilities, one by one.
And if programmers - spurred by a bounty or otherwise - do manage to recreate the imperfections of the human senses and mind, we would have truly created natural artificial intelligence (as against artificial artificial intelligence).
And then, spam will be the least of our problems.
Apr 27, 2009
Iqbal Mohammed is Head of Innovation & Strategy at a digital innovation agency serving the DACH and wider European markets. He is the winner of the WPP Atticus Award for Best Original Published Writing in Marketing & Communication.
You can reach him via email or Twitter.
// Subscribe to blog updates via RSS or email. //