Gordon Smith

New technology can ascertain your political views by the look on your face

After a recent study displayed the ability technology has to guess our sexual proclivity and political choices, I wonder if we’re too quick to automatically fear these advancements.

 

 

Technology: it’s the all-seeing, all-knowing beast that almost certainly will one day control your life – if it isn’t already.

No, not in a “robots will destroy us all” kind of way, or a “but what is the cloud, though?” style protracted rant. We’re talking more in a “there is no hiding your secrets from the unblinking eye of the future” sense, with a little bit of artificial intelligence-based paranoia to boot.

Well, okay, maybe that’s a bit extreme, but technology is very quickly getting to the stage where it genuinely can tell more about a person from a glance than your ordinary, non-cybernetic self ever could.

That’s not necessarily a bad thing, by any means. In fact, depending on how the concept is explored, it could prove to be an extremely valuable prospect: imagine being able to diagnose certain mental illnesses before they strike, or how AI could finally see an end to innocent people being wrongly convicted.

Or, indeed, the inverse: criminals that otherwise would have taken their crimes to their graves would find themselves entirely unable to obfuscate their deeds, the primitive lie detector superseded by the power of computers.

At least, maybe someday.

You’ve probably heard the news that a study from Stanford University found an AI was able to tell if someone was gay from a photograph with impeccable accuracy: 81 percent of the time for men, and 71 percent for women.

Comparatively speaking, human participants were only able to discern 61 and 54 percent of the two, respectively.

Of course, all of this hinges upon the classic “gay face”, and all the questionable stereotyping that entails.

But that’s not the real takeaway of this study – it’s the precedent that matters.

The researchers behind the study believe a similar AI system could be developed in future to determine other traits, including a person’s political leanings, or their IQ. All it takes is the right type of data.

It’s not an entirely unbelievable proposition, either. Researchers have previously trained an AI to detect cancerous lesions.

More significantly, a profiling system using Facebook data, relying upon the information in a person’s profile to “model” their personality, was created by the same researchers behind the sexuality detecting system, and the Trump campaign used a similar model during last year’s campaign to target prospect voters.

 

While it is always easy to play the “technology is scary” card, and to let paranoia get the best of us, our reluctance to accept the role artificial intelligence will play in our species’ future serves only to hinder us.

 

So how is it that machines can know so much more about a person than any human being ever could?

The answer is simple: they see more, and in more detail.

Take for example Google: copy in almost any image, and Google can tell you where that photo was taken. Show that same photo to even the most experienced of photographer, and they’ll struggle to even come close.

Machines – and more particularly, artificial intelligence – represent a great leap in human ingenuity, and present opportunities that mankind simply could not have previously comprehended.

While it is always easy to play the “technology is scary” card, and to let paranoia get the best of us, our reluctance to accept the role artificial intelligence will play in our species’ future serves only to hinder us.

After all, the more we avoid embracing AI, the more we avoid embracing its applications.

Yes, there are privacy concerns. But that’s not a new concept by any means, and arguably not something remarkably more prevalent in the era of self-learning machinery.

If anything, you could probably say more intelligent technology is all the more safe than our current, one-way method of interacting with machines.

Currently, your only real defence against data theft is a strong password. A machine that is able to successfully and consistently identify its user is inherently safer than this – your current computer doesn’t know who’s typing the password in, just that the password is correct.

AI that can tell you if someone is gay or straight is a novelty at best, and a safety risk at worst.

AI that can tell you whether or not you need to head to the doctors, and quick? That’s certainly something worth embracing.

At the end of the day, we will never come close to our machines; be it in terms of intelligence, or sheer volume of information held.

That’s not something to be ashamed of, or something worthy of fear. No, that’s something to be deeply proud of.

We as a species have the tools required to create machines that far and away outrank us in ability, and more than that, we have the tools required to make those machines work to our advantage.

Yes, change is scary. But the sooner we embrace it, the sooner we can master it.

 

Gordon Smith

Journalist by day, cunning linguist by night. A passion for politics, hypnotically involved in human rights. An Australian born with a Japanese tongue, hoping to hold the big wigs in government to account.

Related posts

Top