Bots vs. Brains

Feburary 28, 2018
By
Martin Zizi

In a former post, I asked a very simple question: What would prevent an all-knowing AI bot to usurp your digital identity? The answer is even simpler: Nothing.


In the recognition/identification/authentication world, we have been exposed to many technologies. Some fascinating like 3D facial recognition, some easy and reliable like vocal recognition, some too easy to hack, some too easy to spoof, and then some… the body is the password is the mantra.


Until the recent past, no machine could spoof a voice perfectly enough to fool another machine analyzing the same vocal input. The second machine would always recognize that the initial voice was machine-generated. One year ago this barrier was broken. Hence my simple question.


When one wants to think about it for a second, an AI could be the ultimate spoofer. It was with this question in mind, that - at Aerendir - we developed a technology based on proprioceptive signals that we could capture using plain phone sensors. Unless the super AI bot can emulate the complexity of the human brain (understand: to modelize the millions of neural networks in a brain and using this knowledge to generate a ‘fake’ signal), it will not be able to spoof a brain-based identity system.


So our brains may end up becoming the best biometry to be recognized across the IoT and, at the same time, may offer us the best protection for our own safety. The brain networks are so complex that they can generate stable patterns worth one million character-long passwords. One can hence at the same time: being authenticated, being encrypted and being recognized as totally different from a bot – even the best ones.


An additional afterthought: brains could prevent election trolls too!