The U.K. government will devote £100 million ($126 million USD) to fostering AI hardware development and shoring up possible computer chip shortages, according to a report from The Telegraph. The public sector organizations UK Research and Innovation and the Department for Science, Innovation and Technology are leading this effort.
Semiconductor chips have been in high demand and in limited supply because of the COVID-19 pandemic, stockpiling and other factors. In this economic climate, countries are trying to position themselves as stable havens for chipmakers to shore up supply against another shortage in the possible AI-powered future.
The U.K. is looking for semiconductor chips from these companies because they can be used to power generative AI models.
Jump to:
The £100 million allotment under Prime Minister Rishi Sunak will be used for chips suited for generative AI models, according to the Telegraph. In particular, the allotment asks for 5,000 GPUs from NVIDIA. The goal is to create a nationally-funded AI resource similar to the National Artificial Intelligence Research Resource Task Force in the U.S.
The U.K. also plans to put £900 million toward developing compute capacity, including an exascale supercomputer.
In March, the U.K. government pledged £3.5 billion to tech and innovation broadly, including about £1 billion of government funding for supercomputing and AI.
The £100 million investment is part of the global competition to be at the forefront of innovation in the relatively new AI sector. The U.K., which produces 0.5% of global semiconductor sales, is scrambling to compete with the European Union, U.S. and China, a U.K. government insider told The Guardian.
SEE: Most companies using generative AI task it with code development, content creation and analytics. (TechRepublic)
The U.S. invested $52 billion into its semiconductor industry with the CHIPS Act, which was signed into law on Tuesday, August 9, 2023. The European Union offered €43 billion ($46 billion USD) in subsidies with its European Chips Act, which was adopted on July 25, 2023.
The EU documentation points out that semiconductor chips have been at a premium and difficult to find for the last year. The acquisition of semiconductors depends on a small number of companies that produce them. This limited availability leads to competition between businesses and the countries in which those organizations reside.
Plus, the U.S. and China continue to compete to dominate the semiconductor industry, with the Biden administration restricting some U.S. investment in Chinese semiconductors, and China declaring U.S.-based manufacturer Micron‘s chips a security risk.
While the U.K. tries to establish itself as a home for the AI sector, it is contending with ongoing competition for the GPU chips — a category of one type of chip that typically falls under semiconductor manufacturing — that underlie heavy-duty generative AI models.
Today, the AI industry contributes £3.7 billion to the U.K. economy and employs 50,000 people.
Aside from plans to secure hardware from NVIDIA, AMD and Intel, how the U.K.’s mix of AI investments and regulations will impact business depends a lot on whether your organization is an enterprise, a startup or something in between. AI regulation is a hot topic in the US and the U.K.
SEE: The Ada Lovelace Institute recently asked the U.K. government to firm up AI regulation proposals (TechRepublic)
Salesforce has committed to doing business in the U.K., with a $4 billion investment over the next five years announced in June 2023 and a stated interest in regulation that encourages, not stifles, innovation.
“A clear pro-innovation regulatory framework that compels safe and responsible use of AI is vital, and Salesforce is fully focused on bringing secure, trusted, enterprise ready generative AI to UK businesses,” said Zahra Bahrololoumi, chief executive officer of Salesforce UKI, in a press release.
“The U.K. clearly wants to position itself as more pro-innovation than the EU when it comes to technology, and especially AI,” said Andrew Gamino-Cheong, co-founder and chief technology officer of responsible AI governance platform Trustible, in an email to TechRepublic. “The EU’s AI Act doesn’t provide a lot of ‘carrot’ for good AI, mostly a ‘stick.’”
“A lot of the U.S. AI space is focusing on the U.S. market itself, and so there’s a huge opportunity for other non-U.S. AI players to emerge and cater towards the broader global economy,” Gamino-Cheong said. “The Chinese AI regulations are incompatible with the standards most other countries have, and so the non-U.S. and EU markets for AI are relatively underserved, and the U.K.’s deep connections to the international finance world can help it become an AI powerhouse serving those markets.”
More information about the UK’s investment in AI innovation is expected to be available when the U.K. holds an AI Safety Summit this fall, likely in November, to try to hash out standards for safe use of artificial intelligence, particularly frontier solutions, meaning future-looking machine learning models. As of August 15, a formal date for the summit had not been set, several embassy officials told Politico.
“As AI rapidly evolves, we need a global approach that seizes the opportunities that AI poses while grasping the challenges and minimizing the risks,” said Foreign Secretary James Cleverly in a press release about the summit.
