Allan Nixon: Sunak must make Whitehall fit for purpose if Britain is to compete on AI | Conservative Home

Allan Nixon is Head of Science and Technology at centre-right think tank Onward, and a former Downing Street advisor.

In two dozen tall cabinets in Edinburgh sits Archer2, the UK’s most powerful supercomputer. Unfortunately, it’s not quite powerful enough – the world’s best machine is 56 times faster.

For Britain to have any chance of competing with other countries on AI, we need an urgent upgrade. Thankfully, Jeremy Hunt announced in the Spring Budget that the Government would invest nearly a billion pounds in a brand new exascale supercomputer.

Whitehall’s response? “Computer says no.” Instead of seizing the initiative, the Treasury is putting this novel procurement through the same bureaucratic business case process that they would use for a new motorway or housing development. The process is likely to take years, when it should take months if not weeks.

In a host of other areas where the Government needs to keep up in the global tech race, Whitehall is way off the pace. While R&D investment is rising, Britain still sits outside the top ten OECD countries in terms of science spend as a share of GDP.

The Government says it is backing five frontier technologies. But there’s no plan for one (bioengineering), too little funding for another (semiconductors), and a third, AI, is underpinned by a 2021 strategy that now looks woefully outdated. The National Science and Technology Council – created to drive cross-Whitehall progress – has barely met this year, despite an explicit commitment to it meeting monthly.

The Government has also done little to improve its relationship with regulators and the science community. It last provided Ofcom with a “Statement of Strategic Priorities” in 2019, when Theresa May was in Number 10; no wonder we’re not sufficiently innovative when it comes to telecoms.

And there is huge disparity in funding approval timescales across the seven research councils that dispense billions in R&D funding – from a 76 days average for the Economic and Social Research Council to 230 for the Medical Research Council.

Rishi Sunak has already taken some big steps to back science. Next week marks six months since the creation of the Department for Science, Innovation, and Technology, which has undeniably energised the science superpower agenda. Innovative approaches like the creation of the Foundation Model Taskforce – modelled on the hugely successful Covid Vaccines Taskforce, and chaired by industry expert Ian Hogarth – have shown this prime minister is willing to do things differently.

But it’s not enough. Other nations are placing the full power of the state behind their science and tech goals, often in direct competition with the UK. China has been pursuing its “Made in China 2025” plan for almost a decade. Earlier this year a single Chinese company, Bytedance, put in an order for computer chips equal to the UK’s entire semiconductor budget over the next ten years.

The US has its £41 billion CHIPS and Science Act, whilst Germany is pursuing a £34 billion “Industry 4.0” strategy. The UK meanwhile continues to be nervous about even uttering the words “industrial strategy” – despite half-heartedly pursuing one half-heartedly (including the welcome articulation of the five “priority technologies” where we have a comparative edge).

A new report by Onward sets out a plan to rewire Whitehall so the Government can realise the potential of science and technology to boost the nation’s productivity, transform public services, and compete internationally in a world increasingly driven by pursuit of technology dominance. The proposals have been backed by ten former science and digital ministers.

Four principles should guide this next stage of reform. First, centrality: Sunak has to put science and technology at the core of his agenda, with sufficient investment and urgency.

Second, choices: government must recognise and embrace trade-offs in how it wants to support science and which areas it wants to prioritise. Third, coordination: Whitehall departments need to be completely aligned, as well as regulators and science bodies like UKRI. Fourth, consistency: government must stick to its plans.

The first step to show this government is serious about science is boosting investment. Nine OECD countries, including Germany, the US and Denmark, already spend over three per cent of GDP on R&D – a science superpower should be pushing to the front of this pack. To fix this, ministers should update their target to 3.5 per cent of GDP, with an emphasis on boosting private sector R&D spend.

The second step is to double down on Whitehall reforms. A new Technology Futures Unit, modelled on Singapore’s Centre for Strategic Futures, should be established to sharpen strategic decisions. Labs and other science infrastructure should be designated as Nationally Significant Infrastructure Projects to accelerate development.

Responsibility for universities should be transferred from the Department for Education to DSIT to better coordinate the science agenda. The Government should also instigate an urgent review with UKRI to improve timelines for funding approvals by research councils, trialling approaches like lottery allocation, inducement prizes, and equity stakes.

And then there’s the Treasury. Massive spending decisions are being made in siloes, without regard for how big investments in technology might generate strategic advantage internationally, which risks limiting our science superpower ambitions. A 2018 survey of civil servants who had contact with the Treasury’s spending teams found that just 42 per cent agreed that HMT considered the long-term impact of funding decisions.

The Treasury also lacks expertise to evaluate science and technology projects: according to Institute for Government analysis, in 2020 there were no science and engineering specialists in HMT compared to just under 1,000 in BEIS.

DSIT must be unleashed from Treasury bureaucracy, and have a unique role carved out in Whitehall. Major Projects in the Science Department should be exempt from HMT sign-off to allow ministers to move more swiftly.

The Government should also introduce single business case approvals for DSIT agencies, similar to how the Advanced Research and Invention Agency (ARIA) was formed, free from micromanagement by Treasury officials with little science and technology expertise; responsibility for evaluating DSIT spend should shift to experts, not Treasury generalists, to support the greater spending freedoms for DSIT outlined above.

The Government has shown that it is willing to make big bets in aid of its science superpower ambitions – it cannot now rest. There is a chance for the British state to be uniquely prepared for the tech revolution we are beginning to experience. Ministers must seize it.