[MATCH] Your Face Predicts Your Behavior?

I have heard that, despite the hooplah, polygraphs are quackery, and I have no reason to believe this is not quackery. I have no idea how the lie detector's supporters got that statistic. The lie detector is winning by intimidation, pure and simple. In any case, do we even need Faception's device? Can't we just look at people and tell whether they are lying? Aren't we 100% accurate already?

One way to feel how an iPhone sees more than meets the eyes it the app Cardiio. It can count your heart beat because each heart beat causes a slight blush in the face. I tried it and it is quite accurate. Scary?

I see no mention of compensation for local genetic pools or cultural face clues as part of it's recognition; nor do I see it mentioned that it can spot the subtle multi-racial multi-cultural face shapes and expressions. Thus, there seem to ME to be a built-in bias for specific racial stereotypes --- and, My God!, haven't we moved a tad more towards nurture than nature? I mean, haven't we moved just a tiny TAD closer to learned behaviors than genetic determinism?

Things are often FAR more complicated than they seem. Studies now show that your behavior *MAY* be partly determined by your Grand parents diet while your parents were pregnant, or what type of environment your Grand- (and possibly Great-Grand-)parents were raised in.

I suspect it will utterly fail when confronted by todays multi-racial-ethnic-cultural pluralistic make-up of Humankind.

Sounds like pure hokum to me. Where's any independent research to support the company's 80% accuracy claim?

Didn't an Israeli company come out with a Voice Stress Analyzer several years ago? It was supposed to detect stress when a subject lied when responding to questions. I wonder how that worked out in real life?

The Israelis certainly have powerful motivations to find some way----to find ANY way----to keep dangerous people from entering or remaining in their country. Maybe face recognition will tag some persons and they can then be followed up by being voice stress tested?

By the way, Bob, I think your abbreviation could be made more accurate if it were "WashPoop."

Even if true, 80% accuracy is nothing to boast about. It means that out of 1,000 suspects fully 200 innocent people -- 1 out of 5 -- will be falsely accused.

Of course we don't know what the actual accuracy is. 80% is Faception's own figure, but we're talking advertising English, another term for exaggeration.

The worst prospect is that police departments or airlines put Faception to use. Because it's software the program, hailed as high-tech and digital, will acquire an undeserved panache and the implementers will believe in it.

Bold claims and easily put to the test.

I could probably hit 80% (50%?) just by using the Hollywood method - the bad guys have mustaches.

Sounds like today's version of phrenology. So, based on my face, I can't help being what I am. I'm guessing that they are accepting investors, and in a year or so there will be a little blurb somewhere about this company folding.

IAAL. This is a transparent hoax. The idea that there can be a "criminal face" goes back to phrenology, and none of it has ever come even remotely close to working. Nothing has ever been legally acceptable.

Bob: Can you please send the link to Snopes? This needs to be swatted flat.

Their website gives no explanation for what they base their behavioral predictions upon. They only cite studies which found (surprise!) that certain facial features seem to be more sexually attractive...except when they aren't. Faception isn't even claiming access to some database of mug shots to match criminality to faces. That would be flawed anyway as it is would be limited in scope, contain many unrepresentative expressions, and it would also include everyone who was actually innocent (whether or not they were ever able to establish that). And does there even exist a mug shot database of extroverts or bingo players, let alone a set of personality traits that could begin to separate those people from others who have similar "genetic" qualities but who excel in very different pursuits? (Unfortunately, Facebook's facial recognition already may be trying to compile such data.)

We all know people we grew up with who went on to lead lives we never would have predicted. Some of the really "bad" kids I knew became cops, while some kids who were really "into rules" became organized criminals. There is a thin line between them, including in appearances that Faception wants to be paid for pointing to. Will we really want to stop entrusting police with our safety, perhaps firing huge numbers of them, when some algorithm classifies them as looking like the criminals they deal with? This is digital snake oil being sold to frightened people who have not stopped to do critical thinking for more than 30 seconds. You can tell the premise has been bought when someone quickly starts selecting scenarios where this will save us from evildoers or bad employees.

There are literally millions of targets for a terrorist to strike at. It has to be pointed out yet again that the only way to stop hordes of potential threats is to remove their motivation. There was no motivation to carry out suicidal slaughter in the not-so-distant past but nation states since then, including ours, have callously meddled in (and exploited) other cultures, oblivious to the ideological fanaticism this stirs up. So terrorism will continue unless we all recognize this as the ultimate cause and reconsider our attitudes to those that are different from ourselves.

What absolute rubbish to say your personality is in your DNA. DNA can be changed by RNA which is changed by environment.

While reading this article I, as an Italian, couldn't help but thinking to a modern incarnation of the work of scientist Cesare Lombroso who is considered as the father of criminology (see https://en.wikipedia.org/wiki/Cesare_Lombroso).
Many of his theories however are today commonly rejected and considered as pseudoscience.

Faception does not need to reach reliability of about 99% to become useful. It can be used as an auxiliary tool for indicate who should be submitted to further examination.

For example, it is impossible to thoroughly examine all the people that enter an airport terminal. Technologies like Faception can be used to alert security personnel that a particular face highly complies with known features of terrorists. The documents and suitcase of that person would than me examinewd more thouroughly than those of other people.

Evidently, Faception and similar technologies cannot be used in one particular country in which PROFILING is considered taboo. Even though thousands of people were murdered in that country by Muslim terrorists, the president would not even admit the existence of Muslim Terrorism. It may take another terrorist attack of the “9/11” dimensions to change this situation.

Don't forget Sherlock Holmes and his "EAR" list of shaps to ID the bad boys as well!!

New Age mumbo-jumbo, not real science.

Further to Howard L's statement that, "It means that out of 1,000 suspects fully 200 innocent people -- 1 out of 5 -- will be falsely accused": How many real terrorists would be given a pass by this system? Besides false positives we need to take into account false negatives.

Sounds like something "BIG BROTHER" would come up with and try to justify using !!!

A "criminal' to one culture is a 'hero" to another. How would George Washington pan out? To the Torries of 18th century England, he must certainly have rated somewhere on the "terrorist" / "criminal" scale, but here in the USA, this description would hardly befit one of the country's "founding fathers".....right?!

Is this why passport photos now have the requirement "no smiling?"