In early April, workers at the multinational tech company DXC “walked off the job and onto the picket line” in what their union have claimed as the first strike in the Australian software industry. It couldn’t have come at a more critical time.

The action was the culmination of a long-running dispute. A small proportion of the DXC workforce were employed on old contacts that had been signed prior to a 2010 change in industrial relations law. The law raised standards for pay and conditions, but a legal loophole allowed existing “zombie” contracts to persist, leaving many working on worse terms than would be legal on new contracts. The government set a deadline in late 2023: unless companies successfully applied for an extension, the zombie contracts would automatically expire.

Months before the deadline, DXC management offered those workers a new contract. In their haste, they made two clumsy mistakes. One was to offer a small bonus if their employees agreed to sign quickly. This immediately aroused suspicion. As one of those DXC employees told me, “no one gives you money for nothing. There’s something going on here”. They weren’t wrong. When that employee looked at the details, he found that a cut in stand-by rates would leave him would worse off by $13,000 a year. The rates of many other workers was set to stay the same as it had been for the previous five years — given inflation, an effective pay cut.

The other mistake DXC made was to leave all recipients’ email addresses visible on the email they sent out. Effectively, they had just handed over an organising list. A few union members set up a communications channel outside corporate control, and reached out to everyone affected.

The company dragged its feet, insisting on dealing with the union’s proposals piecemeal, leaving long gaps in-between meetings. They refused to send counter-proposals ahead of time, and sent negotiators who lacked the authority to agree terms. They bundled other workers into the negotiations in an apparent attempt to reduce the proportion of union members among the cohort. After two years of slow-motion bargaining, DXC offered only meagre increases to some of the cohort, and none at all to the majority.

It was at that point the workers voted to strike.

Many of DXC’s tactics will be familiar to workers who have taken part in collective bargaining. What’s unusual about the strike is where it happened. The software industry has typically had low union density. I’ve been a software developer for two decades, and for most of that time I wasn’t even aware that there was a union I could join.

Ten years ago, in a piece for Overland, I noted the “lack of unionisation or any kind of collective workplace action by software developers, despite our frequently poor working conditions”. One reason for this is high turnover. Software workers tend to switch jobs frequently. It’s harder to organise when colleagues are constantly coming and going. There are also fewer social opportunities to discuss matters of collective interest. Interactions at work are increasingly confined to corporate communication channels — videochat and instant messaging — rather than face-to-face. Under those circumstances, people are understandably less willing to speak frankly.

Face-to-face, though, it’s different. Workmates are curious about the union. Recent upheavals in the industry have given them a reason to think about organising. Massive layoffs in the USA and across the world, including in Australia, have been presented as the result of AI efficiencies, but the true cause appears to be often more to be found in the prediction than the actuality of AI’s effects on the industry.

The nature of the work has radically changed. A year ago, coding was my core work activity. Now I barely write any code myself. I tell an AI agent what to do. I review the code it writes. I point out what needs changing. This way of working seems to get things done quicker. But has also drained the work of job satisfaction. The workday becomes exhaustingly fragmented. Software written in this way can be harder to maintain in the long run. Coding is a skill that is sustained by practice. Like other skills, it atrophies with disuse.

This “deskilling” is not merely a side effect of the change. It’s the whole point. What we are witnessing is the privatisation of the general intellect. The vast quantity of land, water, power, concrete, steel, silicon and rare minerals currently being poured into the construction of giant neural networks and data centres are not the only resources currently being consumed by the tech industry. The most important resources are knowledge, skill and the practical combination of the two: knowhow. The giant bet that global capital is taking on AI investment is predicated on replacing intellectual work with a profitable rental service.

The current upheaval is sometimes called the “AI revolution”. Both parts of this phrase are misleading. It’s not a revolution. It brings the abstraction and concentration that were already essential features of capitalism to a new height of intensity. Silico-capitalism is not a break from petro-capitalism, but continuous with it. Countless hours of work by endless hands are congealed and concentrated in capital. as Marx puts it, “capital is dead labour, that, vampire-like, only lives by sucking living labour”. Likewise, the dark, stinking fluid that fuels the engines of the global economy is a concentration of long-dead prehistoric plants and animals: many tonnes of ancient life per litre of petrol. And now, the zombie apocalypse is upon us, but instead of shambling corpses we are being overrun by a flood of undead slop, formed from the accumulation and regurgitation of human discourse and thought. The marketing term “artificial intelligence” implies a new, sui generis form of thought, but this is a sham. It feeds on our words, our ideas, our conversations, our books. It’s theft, and more than that. Class warfare of a particular character. Intellectual necromancy.

How the current disjuncture in political economy plays out will have profound effects on the future. Like other imperial landgrabs, the mass plunder perpetrated by tech giants’ has been dressed up as the inevitable unfolding of history. We should reject these narratives. The future is not yet written.

Neoliberal globalisation used to be touted as the world’s uncontestable manifest destiny. For decades, it seemed that there was no alternative. But history refused to end. New forms of nationalism have sent globalisation into retreat. Regimes are inevitable until they aren’t. The USA’s global dominance in technology is underpinned by its economic centrality and military reach, but it is now being undermined by the chaotic aggression of its government in trade, war and foreign policy. Where once only rivals such as China maintained their distance, European and East Asian allies of the US are now also scrambling to build their own platforms and escape dependence. Other countries may choose to follow suit. Procurement and investment decisions that were once judged merely as value propositions have become fraught with geopolitical risk. The habit of national governments to genuflect to Washington is being met by backlash from the general public.

This cuts across party and ideology. The US-Israeli war on Iran is too toxic in public opinion for even the far-right government of Italy to allow use of its bases. Meanwhile, tech companies directly complicit in Israel’s genocide in Gaza are being granted access to intimate data on the populations of countries across the world. The more the UK public learn about the unlimited access their government has granted US company Palantir to ransack databases of citizen’s personal data, the less they like it.

On these issues, the vast apparatus of rightwing media devotes considerable energy to obfuscating the very common sense they loudly profess to champion. Workers seeking to counter this message do not have billionaires to bankroll them. What we have is what we have always had: each other.

It would suit the interests of our would-be techno-fedual overlords for each of us to simply fret, individually and impotently, about the situation, while the economic foundations of our lives are hollowed out beneath us. If we aren’t happy with this, we should instead act in our own interest: get connected and organise.

One aspect of this is political: to argue, as the Australian Council for Trade Unions has been doing, for sovereign tech capacity, and to make technology a servant rather than a master. Another, more fundamental aspect is building collective power at the workplace. Only on this basis can organisations such as the ACTU or Unions NSW mount pressure for policy changes. It is also the means by which we can directly promote our interests at work.

That’s why the DXC workers’ collective action is such an important first step for Australian software workers — even though the outcome is still uncertain. Negotiations over pay and conditions are ongoing. Management appear to be playing a game of attrition. But the company’s key role in large-scale digital infrastructure and public service provision gives workers additional leverage. At this crucial moment, they deserve the support of other workers in this industry. There’s an online petition to sign, but more meaningfully, Australian software engineers can join the union for their industry, Professionals Australia, and get involved. It’s early days for unionisation in Australian tech, but successes from the South Korean game industry and Swedish payment company Klarna to the Alphabet Workers Union in the US show the potential for collective action to make a difference to our lives.

Image: Bob Mical