Few predictions have aged as badly as those made about social media and the internet a decade ago. Newspaper columnists fawned over ‘Twitter revolutions’ in the Middle East, open source breaking down traditional property relations, and ‘Massive Open Online Courses’ replacing outmoded education systems across the world.
That particular era is over. There are no more calls for Mark Zuckerberg to run for president. Today, it is difficult to find anyone who disagrees with the idea that there is something wrong in the way that the internet functions. And whatever specific criticism they have, it can usually be traced to one of the tech giants that now dominate — Facebook, Amazon, Google, and Twitter.
The fashionable contemporary solution to this problem is ‘Web3’, which promises to democratise the internet by decentralising it. Where a fragment of a system exists on each user’s device, the argument goes, no single entity can amass too much power or influence.
In practice, these arguments about decentralisation are exercises in public relations for the boom in popularity of cryptocurrency and non-fungible tokens (NFTs). Viewed in the cold light of day, both are little more than speculative assets for which returns are only likely when investors can encourage more people to get involved in the game. Far from a radical step towards a democratic future, this begins to look a lot more like a classic pyramid scheme.
But the true genius of these new crypto technologies is how they have coded libertarian and pro-market ideas into our understanding of the internet. By promoting the idea that a truly democratic internet would be one governed by individuals, not collectives, and that this would be organised on the basis of private property, they have done more than any recent politician to take the Thatcherite mantra that ‘there’s no such thing as society’ into the twenty-first century.
It’s not a coincidence that some of the more coherent suggestions for how crypto technology might be applied include ‘smart contracts’, which lock you out of the home you’re living in if you miss a rent payment, or a store of wealth that cannot be touched by ‘oppressive’ governments (for example, if they’re asking you to pay tax).
For all of its faults, the hype around Web3 does tap into something real — a widespread discontent with the hoovering up of the internet’s potential by a handful of shareholders and tech CEOs. Big Tech monopolies are a problem, but they are a problem that must be met with collective action, not libertarian fantasies.
Nor will the liberal solution of breaking up Big Tech be enough to tackle the problem. Services like Google are so useful precisely because of its centralisation: the vast amount of data it processes. The fundamental issue socialists have with these platforms isn’t the service they provide, it’s the fact they do it for private profit rather than in the public interest.
The Reality of Monopoly
Campaigns to limit the harm of tech monopolies have focused on the offline impacts on their lowest-paid workers — Facebook cleaners unionising their London office, for instance, or the GMB organising workers in Amazon’s UK warehouses. Because of their warehouses’ notoriously poor working conditions, calls to ‘Boycott Amazon’ circulate periodically on social media.
But the focus on their retail arm misses the more insidious influence of Amazon. Increasingly, the corporation’s influence in our society comes from our dependence on the tech giant’s ‘Amazon Web Services’ (AWS). These services power a more significant chunk of the internet than Google or Facebook but are obscured under the surface of seemingly unrelated services that impact nearly every aspect of our lives.
For example, to get a Covid-19 vaccination, you need to book on an NHS app hosted by AWS. If you want to stream a TV show — on Netflix, Amazon Prime, Disney+, or BBC iPlayer — your video is stored on an AWS database and processed by AWS servers, with recommendations for your next TV binge computed by AWS hosted machine learning algorithms. Your streaming provider pays for every use of each of these services to Amazon. If you clicked through to this article online from a Twitter link, you used an AWS hosted service.
It is effectively impossible to boycott a company like Amazon, so how do we begin to challenge the power of these tech monopolies? For a more democratic and egalitarian alternative, we need a level of funding to build a parallel, state-managed, and nationalised infrastructure on a scale that cannot be bootstrapped or crowdfunded — and you can’t make it out of your garage at the weekends.
Creating infrastructure on a scale that could wrest control of the services that millions of people across Britain use every day from the wealthiest men ever to have lived would require billions of pounds in funding, and economies of scale that cannot be reached through decentralisation or distributed apps. In fact, it requires the opposite: central planning.
Fortunately, this is not as utopian as it might sound. There are already viable, fully costed proposals for widespread government intervention into the infrastructure of the internet. Derided as ‘broadband communism’ by the pundit class, Labour’s 2019 manifesto proposal to roll out free, public, and full-fibre broadband across the country is likely to be seen as visionary in years to come.
Not only is the so-called ‘British Broadband’ feasible, but it’s already supported by the very workers who would be needed to implement it. The proposal has its roots in research supported by the Communica-tion Workers Union (CWU), which has since adopted it as union policy. It has been approved by its members — including BT broadband engineers. They correctly see it as a means to rationalise a chaotic and under-performing sector.
The proposal was conceived as an infra-structure project to address the UK’s threadbare broadband network. Only 10% of households and offices across Britain are connected to full-fibre broadband, compared to 97% in Japan and 98% in South Korea, with 80% of adults surveyed saying they had experienced internet reliability problems in the last year. The coronavirus pandemic highlighted the need for this policy, with teachers estimating at least 10% of their students struggled to access remote learning due to problems with broadband, alongside issues around access to equipment in lower-income households.
‘British Broadband’ could be expanded to develop the infrastructure we need to build an internet owned by the people who use it. Currently, NHS ‘customer’ data and Transport for London’s analysis of real-time usage data are both handled by Amazon. This means our public services are being used for yet another hidden subsidy to the profit margins of tech giants. But that doesn’t have to be the case.
We could — and should — bring existing outsourced government projects and data into public hands. Aside from anything else, this is the only way to make such data collection accountable to the democratic process. But it has other benefits too. For instance, nationalising this infrastructure would allow cost-effective, strategic state-driven investment in twenty-first-century industries, such as artificial intelligence, which require large cloud computing budgets.
One of the most challenging aspects of this proposal is the national element of nationalisation. In practice, a public internet would involve building data centres — buildings full of servers and cooling units. As the internet is global by design, servicing users of ‘British Broadband’ hosted applications from a single data centre based in the UK would mean inefficiently transmitting data packets halfway around the globe. That is certainly a challenge.
The problem could be solved by building data centres on different continents, much like Amazon’s AWS already does, partitioning its services up into ‘regions’ (Ireland and London being eu-west-1 and eu-west-2 respectively). This would certainly increase initial setup costs, but the BBC World Service already provides a model for a British public service achieving global reach through syndication and local partners. There’s no reason that can’t be replicated in this case. If the public sector is to regain some of the ground lost to the market in recent decades, it will need to find ways to operate in global industries.
The Importance of Planning
The importance of a ‘People’s internet’ to socialist politics goes beyond the need to expropriate big tech or curb their power over our lives. The year 2021 was dogged by supply chain issues, causing food, fuel, and consumer goods shortages — and leading to the cost of living squeeze. Unfortunately, these problems are predicted to worsen as the climate crisis unfolds. The market can’t coordinate the resources necessary to overcome them and wouldn’t do it in the public interest even if it could. This problem can only be solved in the long term with central planning.
To some extent, our economy is already planned. A decade ago, during Occupy Wall Street, a paper published by the Swiss Federal Institute of Technology in Zürich attracted considerable media attention. It argued that a network of around 150 ‘superconnected’ companies effectively controlled the global economy. In 2011, this was largely banks — but ten years later, tech giants like Amazon, Facebook, and Google would be at the top of this list.
The internal decisions made by these corporations have a significant impact on our economy. They allocate resources on a vast scale, decide on investment, dictate production, and organise the pay and conditions of workforces across the world. As Leigh Phillips and Michal Rozworski argue in their book The People’s Republic of Walmart, all of this is planning — but none is subject to democratic control or carried out in the public interest. It is planning for profit.
The kind of planning we imagine on the Left looks different — it involves the key decisions that shape our economy being made by a democratically accountable system. To model an entire economy with any degree of accuracy requires a lot of data and computing power. When Chile’s last socialist government under Salvador Allende tried to carry out a democratic model of planning through its ambitious Cybersyn project, it found that it lacked the level of data needed to make this work. Today, we no longer have that problem.
We do have a different problem, however. The data that exists isn’t public, it is owned by private interests. This poses significant issues for any putative socialist government. Take the Mitterrand government of France, for example. After it was elected in 1981, industrialists organised a ‘strike’ to overturn the government’s programme of nationalisations and wealth taxes. If capitalists own the data that allows them to understand how our economy works, these actions will inevitably occur in response to any significant left-wing reforms of the economy.
Whether tech giants would act similarly faced with the possibility of a left-wing programme today is not a hypothetical question. Even a move from Australia’s conservative government to make Google and Facebook pay local news sites for hosting their content led to Google threatening to switch its search engine off for Australian users. Threatening tech giants’ market dominance is a far more aggressive proposal. They are likely to respond accordingly, holding to ransom any socialist government looking to bring the internet into public hands.
But they will only resist this effort because it is so central to the struggle for democracy in this century. If the internet is to be a genuine public service, platforms must be responsive to democratic pressure and big data and artificial intelligence must be mobilised in the public interest. The alternative is a dystopia: a system of surveillance capitalism in which we are monitored and monetised in the interests of a propertied elite.
There is an alternative, and it is feasible. But it cannot be built without first changing who owns the structures on which the internet operates. As long as billionaires own them, the services we use every day will continue to work in their interests. Under public ownership, we could use it to benefit the people instead.