China’s social scoring system and its use of facial recognition to spot criminals in crowds has been well-documented. We’ve tutted about it and speculated about the harm it can do, we’ve worried that these systems will make their way west, and many of us have also – at least privately – considered the potential benefits to such systems.
Of course, there are some benefits, otherwise the systems would not have been implemented in the first place. The idea that we could more effectively fight crime and capture criminals, that we could be rewarded for being ‘better people’, sounds promising – in theory. But the truth is that systems like those which China is ploughing ahead with are a serious concern to the majority of us, believers as we are in the ultimate good of liberal democracy.
We should have learned from the 20th century that authoritarianism doesn’t go well. In China, in particular, Maoist communism led to widespread poverty and economic failure. In the years that followed Mao, the Chinese were freed from the more severe aspects of the regime, whilst censorship and the deeply entrenched police state continued. These changes spelt good things for the country, whose economy gradually regained its health. China is known generally to be thriving economically. However, things are changing again.
A wealth gap is emerging in China, the ghost of Maoism whispers again. As the Carolyn Zhang put it in the New York Times this week, “China is reversing the commonly-held vision of technology as a great democratiser, bringing people freedom and connecting them to the world. In China, it has brought control”.
Fake It Till You Make It
Those famed facial recognition glasses that were used by the police and attracted much media attention earlier this year are a red herring. They don’t work as well as was made out. The target has to be standing still for a while for the recognition system to actually work. Nonetheless, the mere threat of them has been enough to elicit confessions from criminals in questioning.
China uses propaganda a lot, with much of it recently expressing veiled threats about how sophisticated the country’s technology is. One video, entitled ‘Amazing China’, proudly states: “No matter which corner you escape to, we’ll bring you to justice”. Such propaganda leans heavily on the stories of how facial recognition technology is being used to spot wanted criminals in crowds. Indeed, two dozen criminals were apprehended at this year’s Qingdao Beer Festival.
In Xiangyang, a billboard has been erected on a major interchange. The billboard is linked to CCTV cameras combined with a facial recognition system that captures jaywalkers and displays them on the big screen for everyone to see. It also displays people who haven’t paid their debts, for example. It takes up to a week for the system to process and display offenders on the screen, and humans still have to sort through the images to match them to people’s identities. But what the billboard example demonstrates is just the tip of the iceberg of what can be achieved through the use of artificial intelligence in surveillance.
In 2017, China’s public security market was valued at over $80bn (US). There are currently 200 million surveillance cameras in the country, monitoring 1.4bn citizens. This number is projected to reach 300 million by 2020. A company called Eyecool is already passing over 2 million facial images to a police system, which is actually called Skynet (no joke), daily.
Artificial intelligence technologies have the potential to upend the existing paradigm that authoritarianism equals economic stagnation whilst liberal democracy is the only way to ensure sustained economic success. AI enables social control at an unprecedented scale without costing too much to implement or maintain. Crucially, it’s effective – and that’s what makes it so attractive.
Data is the lifeblood of AI, and we are generating more and more of it every day. Whilst we have been grappling with ways to regulate how data is used here in the west, China seems to see this profusion of information as more of an opportunity than a concern. And as a country of strong market power and diplomatic authority, China has the wherewithal to influence technical standards and to suggest to other governments that controlling the internet and citizens at large has many benefits.
At this point, we must address the fact that artificial intelligence can be used not only to deal with crimes and infractions already committed. It also has the unsettlingly ‘Minority Report’-esque ability to predict. A government that does away with controls on data use gives itself the power to draw on data from a wide range of sources, including:
- Tax returns
- Medical records
- Criminal records
- Sexual health clinics
- Bank statements
- Genetic screenings
- Location data
- CCTV monitoring
- Family and friends
That’s not to say that Western governments are likely to do away with data controls, quite the contrary. But we cannot escape the idea that some will be tempted. In the current political climate, particularly in the US, we cannot rule out the possibility of sharp controls of individual freedom to be forthcoming. It may be dramatic to say that the States could be on the cusp of an authoritarian regime, but it’s not completely untrue.
AI not only offers the opportunity to control citizens, but also to direct and control the economy itself. With the technologies already at hand, we can plan and predict market forces as easily as Amazon can predict what you might want to buy next. That, in itself, is a very attractive thought, particularly for enthusiastically capitalist countries.
I am not from the US. But often for Brits, it does feel like we are the 51st state. We pay close attention to what happens across the Atlantic, having learned over the years that what starts over there seems to trickle down to our little island over time. Things in the UK, in my perception at least, seem to happen very much like America, but quieter and more insidiously. But we’re also a country of people who are fierce about our personal freedoms, in some ways more so even than our American cousins. We’re way behind in the AI game, though. Our capabilities compared to the US and China are Fisher Price, and despite all the myriad horrors we do face from our dysfunctional government, I have faith that we, as a people, are smarter than to give ourselves over to “algorithmic governance” wholesale.
That said, these changes will take place incrementally (as will the gradual development of the UK’s AI capabilities). An article published in Foreign Affairs this week references the cognitive science of influence, and reminds us that “salespeople know that getting a potential customer to perform small behaviours can change attitudes to later, bigger requests”.
Knowing that the monitoring of their physical and digital activities is taking place, people will begin to subtly alter their behaviour so as not to be picked up for exhibiting ‘undesired behaviour’. The subtle alterations in behaviour that people begin to perform in order to stay undetected have the effect, cumulatively, of actually changing their attitudes, leading to self-reinforcing habits. Over time, compliance with the new order feels natural – as if it were our own idea to do so.
So what do you think? Could all this happen here? Let us know by answering our poll: