Summarize this content to 2000 words in 6 paragraphs in Arabic Sir Keir Starmer has launched a plan to harness artificial intelligence across the public and private sectors, which he hopes will cut costs and turbocharge UK economic growth.The AI Opportunities Action Plan, written by venture capitalist Matt Clifford, sets out 50 recommendations. All of them have been approved by Starmer, to ensure Britain is “one of the great AI superpowers” and an “AI maker” rather than an “AI taker”.But experts have cautioned there are major obstacles ahead, from access to energy and computing power, to entrenched public concerns about governance of the rapidly evolving technology and profligate use of private data.What are the government’s plans with health data?Starmer has broadly outlined plans to give researchers and AI companies access to public data sets, including anonymised NHS patient data, in an effort to make Britain an attractive place for businesses and to improve productivity in the struggling health service.Ministers intend for NHS data to form part of a new “national data library” and five public data sets are likely to form the first tranche when the project is up and running.The prime minister did not confirm on Monday whether these data sets would be sold for commercial use, and whether work was being undertaken to determine how the data would be priced.He said, however, that while it is important the government maintains “control” of health records, “I don’t think we should have a defensive stance that will inhibit the sorts of breakthroughs that we need”.Nell Thornton, a fellow at the Health Foundation, said that the think-tank’s polling showed that 75 per cent of the public were happy to share at least some of their data to develop AI systems within the NHS.Still, health specialists and privacy campaigners are sceptical the government will ensure the necessary guardrails are in place to protect private health information and the service itself.Saif Abed, a former NHS doctor and expert in cyber security and public health, said he “fears [AI’s] wanton deployment without stringent safeguards will make the NHS vulnerable”.“Who will perform the governance, audit and oversight function on behalf of the NHS of all these suppliers? How will suppliers be penalised for poor behaviour?,” he added.What is the plan for supercomputers?The government confirmed that an £800mn “Exascale” supercomputer at Edinburgh university would not be going ahead. It will instead invest in a new supercomputer that is more specifically focused on AI.It is unclear how advanced Starmer’s new project will be, how much it will cost and when it will get off the ground but he has committed to increasing Britain’s government-owned computing power 20-fold. Analysts argue that “sovereign” compute capacity — where data is processed and stored in the UK — is essential to ensure British companies and researchers have access to reliable and rapid processing power at a reasonable cost, at a time when access to computing power is becoming a geopolitical battleground. Exascale supercomputers — defined as the ability to produce a billion billion operations a second — are being developed by many major economies around the world, including the US and China. Nscale, an AI data centre company, announced plans on Monday to invest £2bn in the UK over the next three years to build sovereign AI computing facilities, to reduce British companies’ dependence on US cloud providers. Karl Havard, chief operating officer of Nscale, said that the government’s renewed focus on AI was a “catalyst to make this move much faster”. Clifford’s proposals call for an increase in computing power that could be made available to academics and AI researchers — the equivalent of 100,000 graphics processing units in government-owned capacity by 2030. But the target falls short of the capacity already available to Big Tech companies. Elon Musk’s xAI recently opened a data centre in Memphis to train the next generation of its AI model that boasts 100,000 GPUs and plans to increase that figure to more than 1mn specialised AI chips in the coming years. Do we have enough energy to power our new data centres for AI?Data centres require electricity to run and keep cool. How much depends on their size, with relatively modest requirements for average sites. But a giant “hyperscale” centre can consume as much electricity annually as 350,000 to 400,000 electric cars, according to the International Energy Agency. In a report in November, the UK’s National Energy System Operator, the body responsible for operating and planning Britain’s electricity system, predicted demand for electricity from data centres in Britain will grow fourfold by 2030. Yet the UK’s electricity system is under strain, with long waits to connect to the electricity grid and high electricity prices. “There are countries that have taken the lead [in data centres] and we are not one of them,” said Kate Mulvany, principal consultant at Cornwall Insight, an energy consultancy. “If we are looking to excel, energy availability and costs will play a part,” she added.The government said today that it would create special “AI growth zones” where projects will have “better access to the energy grid”. The first will be in Culham, Oxfordshire. Luke Alvarez, managing general partner at Hiro Capital, a London-based tech investor, said the UK needs “a lot more energy fast” to meet its AI promises. How long will it take for AI to boost growth in the economy?The prime minister is under intense pressure to convince markets that his policies will boost economic growth in the medium term as he seeks to keep his government’s fiscal plans on track. A sell-off in bonds pushed Britain’s borrowing costs to a 16-year high on Monday.Starmer said “I don’t think it’s going to take five or ten years to double productivity, not for a moment”, adding that he was “absolutely confident that the timeframes we’re talking about are much, much shorter”.However, most data centres that store and process the data which feeds into AI algorithms take years to build, and the government’s plans for its national data library are in their infancy. Additional reporting by Laura Hughes
