The term Luddite is used today as an insult or a way of claiming someone doesn’t understand a new technology or is resisting inevitable progress. The true history of the Luddites was of a mass working class movement resisting how a new tech was used by elites and bosses to extend their control over workers lives and to profit as they drove down wages and conditions. It took the full power of the British state and its military to crush the Luddites in 19th Century England. Today the potential collective power of the working class movement means it is entirely possible for workers and communities to resist any imposed new tech which aims to control, surveil them and drive down their living standards. Moreover, given the dangers to the environment and human creativity posed by this tech, it means we have every right to question, resist and place road blocks on a tech that is being pushed by the few in the interests of the already wealthy and powerful.
Workers find themselves facing and fighting new forms of technology and its applications; from the taxi drivers fight against Uber, to bus drivers resisting new monitoring tech which imposes impossible scheduling targets, to the fight of Covalen workers against Meta, it’s clear that workers will not simply accept new forms of tech as inevitable or just “progress”. At the same time, more campaigns against data centres in Ireland and across the world show the potential coalitions in communities questioning the future offered by Musk, Thiel, Altman and Co. Socialists should support and welcome all these struggles as they point to a growing resistance to the logic and demands of capitalism at a critical historical moment when we totter on the verge of climate collapse.
A huge effort is underway to make us believe that the AI revolution is coming, inevitable and going to change everything. Whether good or bad, we are told that change is coming and there’s little we can do. From the spread of data centres, to any attempt to regulate AI companies, the message is get on board or be run over in the new AI age. AGI ( artificial General Intelligence) is going to happen any day now and will change everything.
While much of this is hype emitted from those with vested interests in keeping the AI bubble inflated and the venture capital rolling in; the impacts of the new tech can be very real on workers’ pay, conditions and quality of life, and on communities energy, water and other resource needs. Its advocates openly promise millions of job losses and a future with intense AI enabled surveillance, control and a promised loss of capacity for human creativity. We are expected to accept the titans of silicon valley’s dystopian vision for all our futures.
Some on the left, “techno optimists” see the technology behind the AI as basically neutral and one which could have a liberatory impact on humanity. Access to various elements of the technology known as AI is seen as key to the future chances of decent livelihoods, sustainable environments and sustainable societies. Harnessing the potential of AI could yield enormous benefits to all of humanity, we are told. Therefore, it is short sighted to be “doomerist” about this technology, by, for example, opposing data centres and the expansion and development of AI. Instead we should be fighting to ensure poorer people and the working class in general have access to AI. We should be debating how to regulate this new tech but not simply oppose it or be a modern version of the Luddites who smashed new textile machines in 19th century Britain in a vain attempt to hold back the industrial revolution.
Socialists should start by looking at “actually existing” AI; its uses, applications and current direction, not some fanciful potential or promised future AI. How AI is being pushed and used now means the only position socialists should adopt is to oppose it with all our energy, (in effect to be modern day Luddites). This doesn’t mean taking a hammer to data centres in individual actions of sabotage; it means supporting and building campaigns that seek to stop data centre spread and the increased application of AI in public services. It means demanding real regulation and seeking to reduce its use of our energy, water etc etc. The key will be when workers fight against its use, either from threatened lay offs, or efforts to extract more productivity from individual workers by bosses or attempts to replace key public services in for example health or education.
The vast streams of funding needed for AI and the various start up companies at its head means that, increasingly, it is Venture Capital funds who dictate what kind of tech gets developed. The essence of capitalist logic – profit seeking, accumulation and the drive to dominate markets – means that the tech and companies that get the investment are those who an elite few venture capitalists (VCs) believe can deliver on their key priorities as quickly as possible. It is not human need, or what is best for humanity and our environment that determines which tech is developed; it’s the priorities of capital.
Isn’t this the same as always? Maybe, but no tech highlights this more than AI. It may well be the case that certain forms of AI have beneficial elements for some of the world; pattern recognition in health care scans, for example, or some elements in predictive decision making, but any such benefits are unintended consequences of the technology’s development. Aside from the control and possible financial gains of AI to VCs, the promise of the tech for them and other backers is their hope that its capacity will undermine organised working class power. Along with increased coercive control over populations and immense surveillance networks, the intrinsic properties of AI promise to copper fasten existing social, racial and class inequalities and structures of power. Whatever “good” uses of AI there may be, it’s essential to understand that this model of development we are seeing; based on massive large language models, cheap data and every increasing demands to scale up computational capacity is linked to the needs of capitalism and major corporations and NOT anything to do with human benefits.
The reality is the hyper-scale data centers that are blowing the carbon budget, polluting whole neighborhoods, and draining scarce resources are not pioneered to advance human well being. More likely, these centres are used to generate slop, house “dark data”, push generative AI or surveil and commodify every human interaction possible.
A far more widespread probable use of the pattern recognition ability is the targeting of minorities in criminal law enforcement or social welfare systems because it is drawing on existing data sets that contain statistical patterns of discrimination in criminal, welfare or health outcomes. Well-documented cases in the US, Britain and Australia highlight these patterns which have given rise to multiple cases of gross injustice, false charges and people having welfare payments stopped because an AI predictive decision-making technology detected a pattern of fraud or criminality that wasn’t actually there.
While we are encouraged to marvel at one positive use of AI, it should not blind us to the much more pervasive use in other areas of surveillance and control. The majority of the globe’s population will not first encounter the wonders of AI as it quickly diagnoses a cancer or other illness; they are more likely to experience it as they are fired or surveilled by their bosses, or see it add to their workload, or indeed as it is used by the state and security forces in their country, or when it determines whether they can access welfare or health supports. For most of humanity their encounter with AI is more likely to be akin to the experience of Palestinians in Gaza or ethnic minorities falling foul of facial recognition systems than a heart warming early preventative diagnosis of a serious illness.
The costs of AI
This year, the four major “hyperscalers” alone (Google, Amazon, Meta, Microsoft) plan to invest $725 billion on data centres and other infrastructure to support the AI revolution. States across the world, including Ireland, are facilitating the AI boom with infrastructural funds, tax breaks and lax regulations; whatever the tech bros demand, they get. The US alone gave major tech companies a $50 billion tax break to aid the roll-out of data centres, while the UK seems determined to hand over essential public services and their data to ghouls like Palantir.
The scale of the investment into AI is absolutely historic, dwarfing anything in capitalism’s history; even the railroad boom or the early industrial revolution. It means in the next decade, the demands for energy of the AI industry will unleash a massive increase in fossil fuel use; every coal mine, oil well and gas field available is set to be pressed into action to fuel this boom. Even retired nuclear sites of near catastrophic accidents like Three mile Island will be brought back into use. This boom in fossil fuel use comes at the very moment when the precise opposite should be happening to give us any hope of avoiding climate catastrophe. It matters not at all whether future data centres find a way to use excess energy to power nearby homes, or recycle their water in better ways; the reality is (regardless of green washing), this boom will help propel Earth into over 2°C of global warming with devastating consequences for global humanity.
The Paris targets, which were, in any case, woefully inadequate to the task of limiting warming and emissions, have effectively been torn up. Major corporations which yesterday professed their allegiance to climate targets are now opening walking away from them, saying AI will sort this out at some stage in the future (or echoing the Trump and far-right charges that it’s all a hoax).
Secondly, this unprecedented turn to this model of AI development will not just use up the earth’s remaining carbon budget, it will also consume vast amounts of other resources, like rare earth minerals, water, agricultural land, etc. In Ireland they are on course to consume 30% of electricity produced by 2030; a recent UN report says by that year globally they will use 3% of all electricity consumed and the equivalent in water of the basic annual needs of the 1.3 billion people of Sub Saharan Africa.
The decision to do this has been made by a tiny fraction of the world’s elite and rich, driven in part by the insane accumulative logic of capital and their own hubris and profit-seeking. The risks, consequences and negative outcomes will be shouldered by the poorest on this planet and the global working class as well as the world’s ecosystems. The rewards, if any, will be harvested by a tiny few of the already obscenely wealthy. Many of these like Musk and Thiel fund and support movements of fascism and racism to both deflect and uphold their right to make these decisions in the first place.
In Ireland, support for AI translates into obsequious bowing to global tech billionaires, higher energy bills for workers, misuse of scarce water resources, threats of electricity blackouts, surging fossil fuel reliance, a huge bill for the new infrastructure needed to service hyperscalers’ data centres, the obscene misuse of generative AI in the sexual abuse of women and children, and a soft to non-existant state regulatory system reluctant to act against the tech bros.
Despite the hype, AI, like any previous leap in technology, will not by itself end poverty, drudgery or inequality. Millions live in poverty not because the technology we have access to is too primitive but because of the class relations of capitalist society. Railways didn’t end drudgery and poverty. Power looms or the spinning Jenny didn’t either. It needed mass movements to organise and fight in previous eras. The technology to hand can be utilised by those movements just as they can and will be used by the elites trying to stop them. No new tech will, on its own, make this easier or more likely.
Revolutionaries need to be at the front of any AI rebellion. We need to confidently say this tech is not a panacea for the ills we face. We should not accept that it will replace workers in mass lay-offs; concretely we should campaign for a zero tolerance policy on AI replacing any health care workers, teachers or mental health professionals. It is not some abstract AI that is replacing workers, it is employers and often extremely profitable employers making a conscious decision to use AI to decrease wages or increase workloads. None of this is inevitable or for the greater good. We need to see the potential in campaigns against data centers to lead to wider and deeper questioning of the capitalist system that gave birth to them. Anti-AI movements such as the school students’ strikes in Australia, or the spontaneous walkouts at University commencement events show a deepening hostility and questioning of the AI future planned by the elites. We don’t need, can’t afford and have no impute to the AI on offer from Bezos, Altman, Musk and Co.
If that makes us Luddites, then it’s a label we should wear proudly.
