George Williamson wants AI the UK can inspect and keep using
Alan Turing Institute head George Williamson has put national resilience at the center of his research priorities. In a new interview, he argues that Britain’s critical services should not depend entirely on foreign AI systems that can be restricted or switched off. He proposes domestic intellectual property and smaller models for startups, alongside work on weather, farming and disinformation.
Artificial Intelligence··Morning
Williamson proposes smaller domestic models
The Alan Turing Institute, the UK’s government-funded AI research body, could help produce smaller models for startups and domestically owned intellectual property, its head George Williamson proposes. He acknowledges that major AI laboratories may have financial resources beyond the institute’s reach. His October 10 interview makes resilience a leading priority for the institute’s research, without announcing a completed model or a new dedicated budget.[1]
Critical services need continued access to AI
Williamson argues that healthcare and infrastructure, as well as defense and national security, require AI that can be examined and controlled. Foreign systems could be restricted or switched off. Political and commercial pressures, he says, may affect access precisely when it is needed. Protecting critical systems from disruption and countering threats amplified by AI form part of his proposed agenda.[1]
Weather, farming and disinformation enter the research agenda
Public-benefit applications identified by Williamson include weather forecasting, farming and monitoring disinformation. He previously led the government communications center developing technology for security agencies. Earlier government demands to focus the institute on defense and security supply the institutional background. Williamson’s resilience approach could also accommodate climate and environmental research, broadening the areas he places within the security agenda.[1]
