Towards a social democratic approach to science and technology

For a social democrat, the purpose of state involvement in science and technology is ultimately the wellbeing of the people. The conventional way of thinking of these fronts is in economic terms, as about growth and productivity with a nod to public services. This is not wrong. It is why an industrial strategy worthy of the name is central. But it is incomplete for three reasons that today give science and technology a greater political salience than in the past.

First, because we can get the wrong kind of growth out of it – for example, growth that accentuates the existing divide between parts of the country that are prospering and parts that are not, which we call ‘left behind’ but which with their precarious, app-enabled gig economies are in fact as ‘new’ as anywhere else.

Second, because the empire of tech (the stack of technologies built on silicon chips) now extends to the media and politics. Media through the technology of social media, the ownership of old media, and the capturing by Google via Search and YouTube of the advertising revenue that would otherwise support old media. Politics via influence in Washington, direct funding and editorial support for the right, and UK lobbying on an unprecedented scale (most obviously with the almost $400 million that has built the pipeline that runs from former Oracle CEO Larry Ellison through the Tony Blair Institute (TBI) to Peter Kyle at the Department for Science, Innovation and Technology [DSIT]). 

Third, because science (along with law, the media, the civil service and academia) should be a buttress of reason and evidence against the novel form of authoritarianism that now threatens us – one that is full of bullshit, lies and caprice bubbling up from the epistemic swamp of the online right.

Because Big Tech is American, because of the position of tech oligarchs within the second Trump administration, and because of the wider importance of the US to the UK, relations with the US are now central to all these issues. Trump is ushering in a new world order, and social democrats both within and beyond the Labour Party have to think about how this reshapes the challenges we face. 

Understanding our antagonist

In opposition, Keir Starmer developed the idea of ‘mission-led’ government as his organising principle. This owed a debt to economists such as Mariana Mazzucato who see a bigger role for the state in the UK economy and ambitious but socially constructive ‘missions’ as a way of making it work. It was a rejection not only of earlier Conservative approaches, but also of New Labour’s approach. This rejection was crystallised in the growth mission where a central plank was an industrial strategy – both a step towards a developmental state in the mould of the US, Japan or Germany and a sharp break with the hands-off approach instigated by Margaret Thatcher and sustained by Blair, Cameron and May. 

In reaction to this, a new alliance was formed with four corners: Tony Blair and like-minded parts of the Labour Party; William Hague and like-minded parts of the Conservative Party; Larry Ellison, the founder of Oracle, and like-minded parts of Silicon Valley; and Adrian Smith, the president of the Royal Society, and like-minded parts of the scientific establishment. Blair and Hague provide the political leadership; Ellison provides the money and the Big Tech agenda; Smith and his successor, Paul Nurse, provide the credibility of the scientific establishment. The presence of the Royal Society in such a factional gambit may seem surprising but one consequence of Starmer’s developmental turn is that science should cease to be primarily an end in itself, as it has been since Thatcher, and become primarily a servant of industry and society.

This alliance came into view at the start of 2023 when i) Blair and Hague published A New National Purpose, a tech-friendly manifesto that has since provided the basis for more than 70 reports from the Tony Blair Institute; and ii) Rishi Sunak created DSIT, providing a locus within Whitehall and the Cabinet for this alliance that has endured despite a change of government. It is now run by a leading Blairite, Peter Kyle, who has appointed one of Hague’s allies, David Willetts, to run its new Regulatory Innovation Office. 

The alliance is a blend of the old, the new and the troubling. The old – persisting with Thatcher’s anti-industrial strategy position, including avoiding spending on near-market research and instead pursuing a pointless science superpowerdom (for example via a 10-year science strategy that stands apart from industrial strategy). This is in essence the proposition that if we spend more on the stuff that hasn’t worked for 40 years, it will start working tomorrow. The new – promoting technology, especially AI, as the solution to many of the public sector’s problems. The troubling – the hook up with Silicon Valley money and consequently both a push to re-orient regulation and government strategy to favour Big Tech and adopting sotto voce right-accelerationist ideology, including by elevating technology into the national purpose of the UK. 

Because it persists with Thatcherite principles that endured through the Blair and Cameron-May governments, and combines these with the agenda of Big Tech, I have called this alliance ‘Thatcherism 3.0’. 

Since Labour was elected, Thatcherism 3.0 has secured a string of wins for Big Tech. As I wrote elsewhere in February:

Within days of the election in July, Peter Kyle had announced a new cross-departmental drive for tech adoption from within his Department for Science, Innovation and Technology. By September, his Cabinet colleague Wes Streeting was promising the same in the NHS.

The new Parliament returned from the conference recess on 7 October and got down to work properly for the first time. Five days later, Johnny Reynolds, the Secretary of State for Business and Trade, made the head of Microsoft UK the chair of the new Industrial Strategy Advisory Council.

On 3 December, Kyle told MPs that the new government’s approach to regulating AI would be as hands-off as the old one’s. Regulation would be limited to giving statutory force to the voluntary, safety-based regime that has already been embraced by Big Tech.

On 17 December, Kyle announced his preference for a regulatory regime in which creators would have to opt-out to avoid giving Big Tech freedom to train their AI on copyrighted works.

On 11 January, Kyle promised “…a very high threshold…” for the kind of social media content that is to be considered harmful under the Online Safety Act.

On 22 January, Marcus Bokkerink, the head of the Competition and Markets Authority was ousted by Reynolds. His place was taken by a former head of Amazon UK.

Six days later, the CMA announced that the result of a three-year investigation into the monopoly of Big Tech cloud services was to be… another investigation.

The core of this activity is institutional penetration, a basic element of the Big Tech playbook outlined recently by the Nobel prize-winning economist Daron Acemoglu.

Later, when tariffs were imminent, the Tony Blair Institute and an old ally, Peter Mandelson, now the British ambassador in Washington, argued for the UK to abandon the Digital Services Tax (which nets the Treasury about £800 million a year) to court favour with Big Tech. 

The rise of the populists has preoccupied us in recent years and Thatcherism 3.0 does not proclaim itself to the world, instead preferring to quietly envelop, redefine and defang the very concept of industrial strategy. Even so, it is strange that such a significant realignment of established political forces should take place with so little comment.

Towards a new industrial strategy

A central problem for social democrats is that Labour’s industrial strategy has been doubly destabilised. 

On the one hand, we now need to re-arm and so the defence industries will have to play a much larger role. This will be costly, but since it means a wide range of new high-precision and high-tech manufacturing tasks is otherwise welcome. Section 3 of the recent AI Action Plan, which commits the UK to developing a sovereign technological capability on security grounds, provides an example of what may be necessary.

On the other hand, the idea of industrial strategy that Labour developed in opposition has evaporated. The pre-election industrial strategy (inspired in part by Joe Biden’s Inflation Reduction Act) was that of a Green New Deal. This was based on the idea that developing new green technologies would deliver a win-win-win. It would speed the energy transition, provide new high-tech manufacturing jobs in ‘left behind’ areas, and win support from green-friendly voters in wealthy cities. That has now gone, for three reasons.

Firstly, the fiscal position is tighter and the £28 billion originally promised to Ed Miliband is now a distant dream. Secondly, squeezed by the cost-of-living crisis and mentored by a Conservative Party that has resiled from its cross-party commitment to Net Zero, voters have become more sceptical of the costs of the green transition (though not the ultimate goal). Thirdly, China has already taken a leading position in major green technologies such as batteries and electric cars, not least by providing huge subsidies to firms in these fields, an act against which there is no effective redress in the new world order. 

The upshot of this is that while Labour is still using the same words that it was before the election – ‘industrial strategy’ – what those words mean is going to have to be very different (beyond the narrow, enduring commitment to the grid power transition). We are confronted by the need to re-engineer the industrial strategy, in terms of both its substance and politics. 

The difficulty of this reformulation is captured by the image on the cover of the green paper on industrial strategy published last year. Wind turbines are the poster child of the Green New Deal in the UK but, on the sub-sectoral analysis proposed in the consultation document, will not feature as a priority for the strategy because the UK has no revealed comparative advantage there (China and Denmark already ate that lunch).

In terms of substance, we need to think about the industrial strategy as much in terms of developing state capacity as in terms of the initial policy mix. The core ideas in the green paper are i) dynamism, i.e. more productive firms displacing less productive firms; ii) meso-level focus, i.e. building up industrial subsectors where the UK has a revealed comparative advantage; and iii) using the levers of government, including public spending and convening, to pursue these aims. This framework should allow the UK to put greater emphasis on diffusing technology, wherever it comes from, into businesses rather than obsessing over producing new science. These ideas are sound – so far as they go. 

One limitation is that they comprise a resolutely hands-off framework and this I think reflects the fact that our last experience as a country of the hands-on work intrinsic to a developmental state was with Harold Wilson. We need to develop the hands-on capacity. Only then will we be in a position to make mature decisions when another crisis like the British Steel one arises. This means institution-building. 

The difficulty then is that we have a generation of politicians who have grown up in the Thatcherite vacuum and for whom deep, career-defining engagement in industrial strategy was never part of the plan. So the institutions can only be built successfully on a change in attitude in Labour’s ranks.

Another nettle not grasped in the green paper is the need to hold onto our most successful firms. We have lost most of our large high-tech firms to overseas buyers over the past 40 years. And the entire ‘science base’ => ‘start-up’ => ‘scale-up’ => ‘mature company providing good jobs in left behind areas’ model is currently broken because we are losing the vast majority of good companies from this pipeline, mostly to the US, before they mature to the point of actually providing the product or service at scale. 

The objective of retaining successful firms in the UK would represent a clear break both with the past. At the policy level, this should naturally lead to a range of novel policies. The most important of these would be to expand the provision of capital to scale-ups so that they don’t want to leave the UK. The way to do this is via venture capital, where the UK is unique in having a large pool of spare capacity in trained VCs with the ability to make sound investment decisions. The fiscal rules have been revised in a way that makes it possible for the government to provide this capital via a fund-of-funds. This would be intrinsically superior as a method of mobilising capital to grants or R&D tax credits (expert and focussed vs inexpert and diffuse) and has the additional and very important benefit of being an investment that should be profitable as opposed to spending. 

This should be the role of a National Wealth Fund (an investment vehicle) standing free from the British Business Bank (a banking vehicle) under the Department of Business and Trade, and armed with £10 billion a year by the end of the first term. Such an empowered National Wealth Fund would: be a clear break with the past; inevitably lean towards tech as the most lucrative venue for investment while being open to all sectors; be a way of mobilising public funds to individual firms without ‘picking winners’; play on distinctively British strengths; be strategically oriented towards the ‘left behind’ regions; genuinely be an investment in Britain’s future; make sense to the bond markets; and have the capacity to shift the dial nationally on productivity and growth. 

Preparing for conflict with Big Tech

Any social democratic approach to tech is bound to be concerned for the health of the UK’s economy and democracy and this will bring it into conflict with Big Tech across four dimensions: ideological, editorial, commercial, and security. Such conflict would be difficult enough with any administration in Washington; today, it is doubly so. In his Munich speech, JD Vance committed to waging ideological warfare against Europe, and warned Europe against regulating the main channel for that warfare – American algorithmic social media. 

The scale, complexity and importance of our relationship with Big Tech speaks to a need to address it in the round rather than piecemeal. Notably, there is a fundamental mysteriousness about the relationship with Big Tech that is being promoted by Thatcherism 3.0. On the one hand, it seems to demand a range of concessions on the part of the UK government (such as appointing an Amazon guy to chair the CMA). On the other, we are never told exactly what the UK is getting in return. 

For all the potential tactical pitfalls, we should therefore be intensely relaxed at the strategic level about conflict with Big Tech. Australia is a useful example. Canberra has legislated twice to clip Big Tech’s wings, obliging it to restrict social media to over-16s and taxing Google News to support the old media it ransacks. Both times the catastrophe-laden warnings from Big Tech came to nothing.

Ideological

The commonalities between the way politics is developing in the US and the UK need to be acknowledged rather than discounted. There is scope for a Trump-like administration in the UK, most obviously under Nigel Farage, and Big Tech has a significant role in this potential.

The ‘techno-optimist’ label given to itself by Thatcherism 3.0 is a veneer for a clutch of darker ideas that circulate in Silicon Valley centred on the right-accelerationism of Nick Land and Curtis Yarvin. This has a contempt for elementary humanism that stems in part from an enthusiasm for replacing humanity with robots and, along the way, treating people like robots – disposable units that are to be retained only so long as they serve an instrumental purpose. A saccharine expression of the same accelerationism can be found in the half-baked ‘manifesto’ drafted by the Silicon Valley VC Marc Andreessen.

Andreessen’s text is widely approved of in Silicon Valley. Elon Musk’s Nazi salute fits in easily with the repudiation of humanism. So does the political orientation of the Palantir corporation, whose founders are vocally backing the worst authoritarian impulses of Musk and Trump. In short, any service provided by Silicon Valley today comes with the risk of it injecting dark enlightenment politics into the UK.

The accelerationism feeds into British politics, and not just via Farage and online trolls. It can also be found in A New National Purpose. This elevates technology to the purpose of the nation, a key accelerationist tenet and lays the basis for Kyle’s cross-government software remit in DSIT in terms that echo the justification of DOGE used by Musk.

Editorial

Francis Fukuyama has framed the power of algorithmic social media as an editorial question. This makes sense because the decision of what to show to whom is perhaps the most basic of all editorial choices. Fukuyama’s characterisation helpfully separates the truly social dimension of social media (chatting with people in your network) from the editorial one, and thus brings algorithmic social media into the same conceptual framework we have for broadcast and newspapers. It is a radically different conception to the narrow harm-based approach to regulation pushed through by Kyle’s predecessor at DSIT, Michelle Donelan, and now championed by Kyle.

Social democrats should therefore call for algorithmic social media in the UK to be regulated by a UK regulator enforcing UK editorial standards. This would undermine its role as a megaphone for MAGA-style influencers and bots – be they American, British or Russian – most obviously because of their readiness to enthusiastically recirculate falsehoods. This is a necessity if we are to avoid determined attempts to upend our democracy – and hence a clear red flag for Musk and Vance. It could also be used to establish a regime for children that is constructive rather than destructive – along the lines of how Douyin in China differs from the TikTok provided to the rest of the world.

Commercial and Security

Big Tech is intrinsically monopolistic and the companies maintain a ‘kill zone’ around the key properties where they either acquire or destroy competitors. This both harms consumers everywhere and prevents competitors from developing, a particular problem for any country other than the US, both for reasons of economy and sovereignty in an era of supply-chain warfare.

Competition law relies on complaints from individual companies harmed by monopolistic practices and has failed to rein these practices in. There is now, or at least was pre-Trump, a determination among the world’s leading regulators, including the CMA in the UK, to tackle the problem. The extent to which the CMA and its new Digital Markets Unit remains on this field of battle today is unclear.

Social democrats must demand a renewed regulatory commitment to competition, including both eliminating kill zones and breaking up monopolies. This in turn will involve reversing out of appointing Big Tech placemen to key positions within the regulatory and strategic agencies. 

Palantir has proposed building ‘an operating system’ for government online services (more back-end than front-end). This speaks to the technological moment we are in. There is a need and scope for coordination at the technical level across government, partly thanks to AI. However, rather give such a contract to Palantir, it makes more sense to use this as an opportunity to expand and develop the in-house government expertise that finally brought sanity and usability to the gov.uk domain after the decades of fragmented and privatised chaos Thatcher ushered in. This is a core capability for government where requirements will endlessly evolve and it makes no sense in principle to outsource it at all, still less to a Silicon Valley firm that will be an enthusiastic authoritarian ally to a British Trump.

This specific example generalises. As the core administrative functions of the state bureaucracy are encoded into software, they become the potential vehicle for an overnight DOGE-style putsch that would bypass resistance from democratically-minded civil servants. Bulwarks against such an eventuality need to be built into the design, procurement and ongoing administration of these systems. This will include denying access to foreign companies that are hostile to democracy and migrating from providers in the US to providers that are not so vulnerable to political forces that are antagonistic to the UK’s democratic foundations. 

As an organising principle, instead of DSIT’s technological solutionism allied to contracting out to Big Tech, the social democratic mandate for technology across government can be embodied in the engineering idea of whole system analysis. For example, the introduction of new technology into the health and social care system only makes sense in the context of a full accounting of its likely effects that doesn’t, for example, overlook the bottlenecks created when partially competent software replaces fully competent staff. 

In other words, technology is welcome, but only when it does in fact improve things. Such an approach has been developed in healthcare by the Royal Academy of Engineering. It could be rolled out more generally by creating a post of Government Chief Engineer in the Cabinet Office and transferring the DSIT cross-departmental remit to them. Again, this naturally entails expanding state technological capacity, and this also creates opportunities for UK start-ups to help meet UK needs – in particular with regards to AI. 

An alternative new world order

If we are not simply to become the victims of the new order being ushered in by Trump then an alternative vision has to be articulated and pursued. Gordon Brown has set out the broad contours of such a programme:

We need a bold, international response that measures up to the scale of the emergency. In the same way that, to his great credit, the prime minister has been building a coalition in defence of Ukraine, we need an economic coalition of the willing: like-minded global leaders who believe that, in an interdependent world, we have to coordinate economic policies across continents if we are to safeguard jobs and living standards. …

All countries that believe in international cooperation should pledge that through a new multilateralism this generation will deliver global solutions to what are now inescapably global problems that cannot be resolved by nation states acting on their own or in bilateral deals alone. … as building blocks of that future, this coalition of the willing should immediately engage in practical cooperation on urgent concerns for which no nation state-only answers are possible – global security, climate, health and humanitarian needs as well as the flow of trade. They should work to modernise the international institutions that deliver them.

This is a vision that is as problematic for China and Russia as it is for the US. It is a bet that the swarm can hold its own against the big beasts. Trade-wise, it is plausible. China accounts for 12 per cent of global exports and the US for 11 per cent, but the next five countries in that list, all natural parts of the coalition of the willing (Germany, UK, France, Netherlands, Japan and Singapore), add up to more than the two combined. For the UK, it is the only alternative to becoming a kind of Bulgaria circa 1960 – poor, unfree and dominated by the ideological invasion that JD Vance signalled in Munich. 

There are four things with which I would round out Brown’s sketch. 

The first is that our understanding of the coalition of the willing should be extended to include all the friends of freedom and democracy that are still to be found in America. Not only do we share common values, but we need each other: the single biggest constraint on Trump’s capacity to advance down his anti-democratic road is US public opinion. 

The second is that unless the coalition of the willing is to be simply NATO-without-the-US, there will be a need for confidence-building measures, in which science has traditionally played a leading part.

The third is a sense of humility in Europeans. Often what we are seeking has been done already, better, in Asia. Where in the world are we going to find the semiconductor hardware expertise that is essential to any kind of 21^st Century security? Non-aligned India has already made good progress in weaning itself off Big Tech with the India Stack of software services.

The fourth would be to include technology and science in the list of urgent concerns for which no nation state-only answers are possible.

Technology

AI is an obvious technological task requiring a coalitional response, the threat being so uncontained, the current position of the coalition being so weak, and the cost being so high. The idea of a CERN for AI is an example of an existing proposal that already fits into Brown’s schema. And the same logic applies to the basic digital infrastructure that we currently rely on, from payments systems to cloud storage and the tools that make data tractable at scale – not to mention the hardware of chips and satellites.

A straightforward way to address these needs would be to develop the EU’s Horizon into a truly coalitional programme of cooperation. However, this cannot be viewed as ‘the answer’ or even the main part of the answer because: i) in a globalised world, success in high-tech domains increasingly falls to those who can mobilise the most capital behind one of the several firms competing to dominate a niche; ii) this is the step at which the US and China are most effective and the rest least effective;  and iii) Horizon, a pre-competitive interpretation of François Mitterrand’s idea of competing technologically with the US through Grands Projets, is by definition of little help. We need to find ways to aggressively mobilise capital behind potentially world-beating technologies. 

Science

Science considered as a whole repeats much of the AI predicament. There is no knowing when some corner of it will turn up something critical and no nation wants to risk being blindsided. Once again, the task is too big for any one country. 

Further, Trump 2’s effect in the US may be more dramatic in science than in technology. The Soviet Union had ideological buy-in from its scientists. China has built its enterprise on a general acceptance of the leading role of the Communist Party. Trump 2, by contrast, lacks such a dominant position in society and has only an antagonism to science that it shows no sign of wanting to leave behind. This hostility is corrosive of the rigour, human imagination, enterprise, investment and camaraderie at the heart of American science. Thus we can expect an acceleration of the ongoing transition to multipolar science, creating opportunities for the coalition.

Conclusion

In the 1960s, Harold Wilson made a determined effort to re-orient the state towards industrial strategy with social purpose, a goal that is naturally central to any social democratic programme. Today, social democrats should once again set their sights on creating a developmental state. 

Back in opposition, Starmer opened the door to such an approach, but we are now almost a year into government with little to show for it on this front. The pro-Big Tech ideologues of Thatcherism 3.0 have succeeding in dominating the government’s agenda, while the expansion of tech’s empire into media and politics, its imbibing of a dark ideology, and its centrality to Trump 2 has dramatically raised the stakes. 

This all speaks to an urgent need for social democrats to face the challenges of technology head-on. This essay has sketched out some broad directions, but going forward we will need to develop new networks, ideas, and institutions through which a concrete social democratic agenda can be formulated, and the shallow hegemony of Thatcherism 3.0 directly challenged.

William Cullerne Bown has been writing about the politics of science and technology for 40 years and was the Labour Agent in Hampstead & Highgate 1987-1992. His writing can be found at https://williamcullernebown.substack.com.