Facial recognition technology is nothing new: casinos and law enforcement have long used it to weed out criminals and black-listed players by matching data-banked images with in-the-flesh humans. But now an Israeli start-up called Faception has developed technology that it claims is better at understanding our personalities than we are ourselves.
Faception recently reportedly signed a contract with “a homeland security agency” (if they mean in the US, we’re pretty sure there’s only the Department of Homeland Security, but then again, we don’t have top clearance, so who knows).
With this new technology, Faception claims it can detect 15 categories of personalities, including terrorists, bingo players, and yes, poker players. And apparently, it’s all pre-determined from our very DNA, so you can stop castigating bingo players, because they’re born that way.
According to the company’s chief executive, Shai Gilboa, our personalities are determined by our DNA and the key to unlocking these secrets is through our facial expressions.
Master of Tells?
Naturally, this statement won’t come as much of a surprise to live poker players, but Gilboa’s claim that his software is better at picking out personality traits than humans ever could might ruffle the feathers of poker-playing masters of this art, such as Phil Hellmuth and Daniel Negreanu.
Hellmuth, of course, is a 14-time WSOP bracelet winner, so is entitled to some bragging rights in this arena. And with almost $32.5 million in live tournament cashes, six WSOP bracelets, and a 2014 induction into the prestigious Poker Hall of Fame, Negreanu has nothing to apologize for in the tell-reading arena, either.
But Faception claims that its 15 built-in “classifiers” allow it to achieve 80 percent accuracy rates when picking out certain personality traits, a win rate that would be daunting at a poker table indeed.
Players Can’t Hide their Poker Faces
To prove the software’s reading abilities, Faception analyzed the faces of 50 amateur poker players before a tournament it had organized.
Using images of each participant and comparing their facial features to a database of professional poker players, the software picked out four players it thought would be the best, based on how well they matched up to the pros. In the end, two of the four players made it to the final three, a result which was seen as a success by Gilboa.
However, as impressive as that result was, a number of issues have been raised by Professor Pedro Domingos of the University of Washington and Princeton University’s Professor Alexander Todorov when it comes to the Faception data.
For Domingos, the possibility of outdated information in the system’s database could skew the results in the wrong direction. Additionally, he suggested that there are possible ethical issues with a piece of software that can potentially say someone has a murderer’s personality.
According to Todorov, physiognomy (the assessment of a person’s character through their outward appearance) is not only unreliable, but outdated.
“Just when we thought that physiognomy ended 100 years ago. Oh, well,” said Todorov.
Regardless, the Faception system appears to have some skills when it comes to picking out poker talent. Perhaps backers will want to check it out before selecting their next tournament horse.