“Who watches the watchers?” asked the first century Roman satirist Juvenal not, as many of us believed, Captain Picard. Juvenal was referring to the difficulty of employing someone to make sure your spouse didn’t cheat. In the millennia since, the phrase has come to mean we should keep our eyes on anyone in power. Roll forward two thousand years and the quote gets another outing: as a warning we have new observers to pay attention to.
I Never Forget a Face
A recent report from tech research company Comparitech ranked the most surveilled cities in the world. Unsurprisingly, China won eight of the spots in the top 10, but coming in at number six, and sporting around 630,000 cameras, was London. The British capital has approximately one recording device for every 14 inhabitants. According to cctv.co.uk: “Anyone going about their business in London will be caught on camera around 300 times per day.” That’s a lot of data.
It is a truth universally acknowledged, that a corporation or government in possession of a fortune in video data, must be in want of a facial recognition algorithm. I expect to see many of them searching for their dream AI at Minds Master Machines, the machine learning conference happening in London at the end of this month. And what’s wrong with that?
Phew, It’s a Good Job We Live in a Stable Democratic State
“You can no longer have anonymity through obscurity,” stated Dr.Joanna Bryson noted AI expert from the University of Bath, “People will be able to find out about you, no matter how unimportant you think you are.”
According to Bryson, in the past few years, data scientists have become extraordinarily good at predicting our future behaviour from information about what we’ve done in the past – and they are going to get even better at it. In our relatively recent world of mass surveillance, as described by the Comparitech report, your destiny will be driven by your history – or even someone else’s. In this new society, governance will matter. Data that could be used to model, or even manipulate, you will not always fall into trustworthy hands. That’s not science fiction and I should know – I write the stuff.
Is surveillance-driven control inevitable? Does the Cambridge Analytica scandal prove the genie is already out of the bottle? Has the horse bolted? Are we just trying to stuff escaped cliches back into the opened box?
Perhaps, but the protesters in Hong Kong clearly don’t believe so. A recent march to oppose the state’s installation of smart lamp-posts equipped with sensors, closed-circuit cameras, and data networks may have ended in police violence but their rebellion continues.
Dumb but Dangerous
The odd thing is, ML algorithms are phenomenally powerful but strangely unintelligent – or, at least, they are not contemplative.
In his new book, “The Book of Why: The New Science of Cause and Effect”, Judea Pearl, winner of the 2011 Turing Award for fundamental contributions to artificial intelligence says, “We live in an era that presumes Big Data to be the solution to all our problems [..] but data are profoundly dumb.”
According to Pearl, Big data combined with ML can definitely be used to predict the future – enough to make you buy stuff or vote for candidates who might not be in your best interests – but machine learning can’t explain your motive. AI can tell us that things are correlated. It can’t tell us why. When Britain’s Vote Leave campaign used data to influence voters in the UK EU referendum, even they could not tell you why people voted as they did – that’s not how ML works. They could, however, tell you that people who saw certain images or read certain sentences would subsequently behave in a particular way.
So surveillance is giving us huge quantities of data, and ML is letting us process it in a very specific way. Not to understand causation – the reason we behave as we do, because AI can’t do that – but instead to examine correlation. Once we might have said, “Tough on crime. Tough on the causes of crime”. We’ll have to replace that with, “Tough on the statistical correlates of crime.” Catchy.
Who’s Zooming Who?
We all know China has a high level of state control but the British government is also surprisingly unfettered. It was once described by former Lord Chancellor Lord Hailsham as an elected dictatorship.
It may be a coincidence that two of the world’s most internally powerful states also have the highest levels of observation of their own citizenry but I suspect not. However, the relationship could, again, be one of correlation rather than causation: the UK’s citizens may just accept weak constitutional checks and balances as well as high levels of observation. In other words, we’re naive fools. But, in a nice, polite way. I wouldn’t worry about it if I were you.
Stepping back, where is this going? In the past decade, China’s internal security spending has increased tenfold and now significantly outstrips their external defence budget. By 2025, cctv.co.uk expects the number of cameras in London will top one million. Apart from the obvious (dazzle makeup and big hats) is there anything we could do about that surveillance, even if we did care?
The Irresistible Lure of the Seemingly-Free Carrot
Surveillance is not all privacy infringement and state control. If it were, we would probably say something. Or maybe we wouldn’t. No, there’s plenty of good stuff mixed in.
Most of us love cameras. We’re desperate for the day we can wander into a Whole Foods store and walk out with a bag of carrots without having to go to the almost superhuman effort of digging out a wallet. I can hardly believe we’ve put up with the monstrous tyranny of purses for so long.
The principle of the slippery slope is often used to introduce new tech or processes that might otherwise seem authoritarian. A year ago, it was suggested to me that children’s hospitals should install facial recognition systems to detect the presence of well-known baby murderers. Who could possibly object to that? The genius of Amazon, however, was spotting that toddler-saving was not required. The thin end of the mass surveillance wedge was mildly more convenient grocery shopping. The figurative carrot was an actual carrot.
What might Amazon use all that new data for? Don’t fash yourself. I’m sure it’ll be fine.
Hang on though, what if something happened to that nice Mr Bezos and Amazon was taken over by someone evil? What then?!! As Dr Bryson points out, perhaps we might want some more laws on the subject. I’m British, though, so I don’t like to mention it.
Anne Currie is the Chief Strategist at Container Solutions and the author of Utopia Five and the Panopticon mass surveillance science fiction series.