Now is the time for a revolution: a revolution in how the forces that will shape our future will themselves be shaped. AI is political, and it is gendered.
The more women speak up and smash the boundaries in work, the more they are told that they are replaceable, as AI is allegedly coming after their jobs. Smart home technology is further enslaving women, and the more the #MeToo movement exposes how women are harassed at every corner, the more the market is inundated with subservient and flirtatious female personal assistants, ready – programmed – to be shouted at.
There is clearly an argument that this is the product of a resentful backlash from men claiming to feel displaced: the more progressivism brings freedom to women, the more AI products and services try to curb it – with AI ready to propel the male fightback. Within these pages I argue that this perceived male displacement is finding a voice in those new populisms around the world that are nurtured, fuelled and legitimized by AI-driven online manipulation. There is a direct and unmissable connection between AI and the political climate around the world, from Viktor Orbán to Donald Trump, by way of Jair Bolsonaro and Rodrigo Duterte. These populisms all have one thing in common. They advocate for a world where women are beneath men.
AI, automation and the digital ecosystem are at the heart of the problem, both as expressions of inequality and as the means to perpetuate it. Automation transforms the economy so that the kind of tax-and-spend policies that social democrats usually rely on become obsolete, making it incumbent on the left to rethink its approaches.
Populism also thrives on the digital ecosystem – and that is thanks in large part to the digital architecture of persuasion, control and oppression that has built up over recent decades. This architecture enables the collection, analysis and manipulation of our data to identify patterns of behaviour and to extrapolate our weaknesses, fears and desires.
The digital ecosystem is booming. Thousands of companies (some famous, some all but unknown) harvest our most intimate secrets, internet searches, online friendships and digital journey. They create profiles of individuals by hacking into our weaknesses.
All this is happening daily; the foundations of our democracies are starting to crumble, and we are not paying enough attention. With individual users having access to personalized news, we are losing the common shared knowledge that once bound us together as citizens. And what can we talk about if shared facts no longer exist?
It is as if we are all the subjects of some global conjuring trick, being distracted by the dazzling light of our shared digital future, and all we are supposed to do is marvel at AI’s remarkable promise. The tech elites present us with a utopian future where AI solves all our problems, reads our minds, detects cancer more accurately and a decade earlier than any human doctor could, and is able to help us with our daily tasks.
Robotics technology will undergo rapid advancements, using artificial general intelligence (AGI), to the extent that some predict that robots will eventually be able to do pretty much anything humans can. But the simple fact is that, so far at least, AI artefacts can only do the tasks they are programmed to do, and nothing else.
While all this (alleged) progress happens, we are risking overlooking what AI can do right now: control, manipulate and remove our autonomy over our thoughts and choices.
In the global race to equip ourselves with the best technology, the trajectory of AI seems to mirror that of nuclear power. The more a nation embraces AI, the better equipped it is to compete internationally. We are in a new arms race, and this race has no end – indeed, no obvious goal in sight. This is a danger to all, not just in the future but in the present, especially when it happens alongside the rise of AI-fuelled populism, which is not merely arresting but reversing social progress along the way.
The power underpinning AI is data. Data, in its current form and scale, constitutes a new form of capital, and its extraction and exploitation lie at the heart of the rise of AI. Data is who we are; what constitutes our intimate being. It is our every online purchase, decision, opinion; every click of our keyboard; increasingly, every movement and sound we make in front of our computer is being turned into a commodity. The harvesting of us as data citizens through control, surveillance and constant monitoring is what makes data turn into capital and accumulation.
But data has a huge flaw, a flaw that is widely ignored or wilfully disguised: data is not neutral. It is inherently political. What to collect and what to disregard is an act of political choice. This means that every single AI artefact – be it a virtual personal assistant or a piece of software diagnosing the presence of polyps – is a political product. It is based on and extrapolated from existing data and therefore reflects the hierarchies and social structures we are immersed in. Data does not simply reflect society as it is, it also embeds those power structures into every corner of our lives by coding them into machines. These machines are increasingly ‘softwaring’ us out: they make decisions for us, shaping our lives and defining what we can and cannot see. They might determine whether we can be hired, whether we can secure a loan from a bank or receive government benefits, or if we should be placed on an at-risk register, with all that that might entail.
I must be clear from the outset: I am not anti-AI. AI has unparalleled potential to transform society for the better in any number of areas. But that AI has the potential to do good is not the point. Like nuclear power, AI can bring enormous opportunities but to do so, it requires a form of authority enshrined in global governance to avoid its terrifying downsides. Like nuclear power, we know that it will affect the politics and geopolitics of the world we live in. It will do so by reshaping labour, not just in terms of our professional lives but also in terms of the meaningfulness of life. These are profound and vitally important issues.
The recommendations of the European High-Level Expert Group on Artificial Intelligence include a suggestion that the focus should be on AI replacing menial jobs.1 This may sound like a reasonable suggestion, but who defines ‘menial’?
The crucial point I am making here is that AI is much more than technology. It is power, and, as such, it has power structures, and these can and do imply, even necessitate, dominance and oppression. All of this we are, perhaps naively, encoding into our systems. The implications are being felt and seen already; for instance in algorithmic racism and chauvinism: those pieces of recruitment software that will only recruit men, or the virtual assistants that, if harassed, will respond to the user with a submissive, coy flirtation,2 instead of insisting that ‘no means no’.
There is no simple solution. The fact that we can intervene in the products doesn’t mean we can resolve the underlying problem. Fixing an algorithm will not answer the question of what, or who, AI is for in the first place.
It is not by chance that one of the tech mantras of this decade is that we need to get women into coding. Awards, programmes, mentoring and special events proliferate, but with little success, as women in AI are still less than 25 per cent of the workforce.3 This is undoubtedly a challenge for our time, but it is only the start of the solution. Professional culture within the tech industry has to change – radically and quickly.
If AI is both power and oppression, we should claim the first, and resist the second. Resisting the second means understanding, not overlooking, what AI can do now. It means reclaiming our personal data, so we can stop the expropriation of our personal autonomy, choice and freedom of thought. There has to be true representation at all levels.
If AI is power, then it needs to be treated as such – and the battles are going to be on company boards, in government, in the media and in international institutions. That is where the law, the demand for accountability, funding and partnerships will be decided and shaped.