In a move to tackle the challenges and opportunities presented by the rapid advancement of AI technology, the United Kingdom government has committed £300 million (US$366 million) to establish a national Artificial Intelligence Research Resource (AIRR). The funding details for AIRR were disclosed during the AI Safety Summit held on November 1-2, 2023, which brought together representatives from 27 countries, including key figures such as United States Vice-President Kamala Harris, European Commission President Ursula von der Leyen, and industry leader Elon Musk.
The international declaration highlights AI risks
One of the summit’s key outcomes was an international declaration signed by the UK, the United States, the European Union, Australia, and China. This declaration raised concerns about the “potential for serious, even catastrophic, harm, deliberate or unintentional” stemming from advanced AI models. It emphasized the need for collective efforts to understand and mitigate risks associated with frontier AI.
Cooperative measures to address AI risks
To address frontier AI risk, the declaration called for international cooperation to:
Identify shared AI safety risks: Collaboratively identify AI safety risks and establish a shared scientific understanding of these risks as AI capabilities continue to evolve.
Develop risk-based policies: Create risk-based policies across countries to ensure safety in the context of these shared risks. This includes transparency, evaluation metrics, safety testing tools, and public sector capability development.
Historic venue and cutting-edge supercomputer
The summit was held at Bletchley Park, renowned for its role in World War II codebreaking and the pioneering work of Alan Turing. Part of the investment includes £225 million to create the Isambard-AI supercomputer at the University of Bristol, set to become one of Europe’s most powerful. Isambard-AI, expected to open next summer at the National Composites Centre, will be ten times more powerful than the UK’s fastest supercomputer.
Crucial applications of Isambard-AI
Isambard-AI will play a vital role in various fields, including:
Robotics: Harnessing AI’s potential for robotics applications.
Big data**: Advancing research in managing and analyzing large datasets.
Climate research: Contributing to climate research efforts.
Drug discovery: Accelerating automated drug discovery processes.
Isambard-AI will also connect with the Dawn supercomputer at the University of Cambridge, forming a powerful AI research cluster.
Frontier AI taskforce and AI safety institute
The UK government’s newly established Frontier AI Taskforce will have priority access to AIRR resources to address advanced AI’s potential national security risks, including bioweapons and cyberattacks. Additionally, the AI Safety Institute will research the safety of frontier AI models and support government policy decisions.
Securing the future with AI
Announcing the investment, Secretary of State for Science, Innovation, and Technology Michelle Donelan emphasized the importance of adopting AI safely and positioning the UK as a world leader in AI safety. The investment is designed to give researchers and scientists the tools they need to delve into AI’s complexities.
Timetable and collaborative efforts
The first phase of Isambard-AI, hosted at the National Composites Centre in Bristol, is expected to be implemented by March 2024. The project is a collaborative effort involving the GW4 group of universities, comprising Bath, Bristol, Cardiff, and Exeter. The entire AIRR program will significantly increase the UK’s AI supercomputing capacity, with Isambard-AI becoming the UK’s most powerful supercomputer.
Tripling the investment
The UK Research and Innovation (UKRI) stated that the investment triples the £100 million outlined in the British government’s 2023 Spring Budget for a dedicated AI Research Resource. When fully operational, Isambard-AI and Dawn will collectively provide at least a 30-fold increase in the UK’s AI supercomputing capacity, enhancing research opportunities for scientists and researchers nationwide.
Global recognition and future summits
The significant investment in supercomputing and hosting the AI summit have been widely praised as a demonstration of global leadership. The UK government has plans for another full-scale AI safety summit to be held in France next year, underscoring its commitment to addressing the challenges and ensuring the safety of AI technology.
Ciaran Martin, a professor at the University of Oxford, praised the government’s initiative, emphasizing the importance of not repeating past mistakes by becoming dependent on technology without adequate security measures. The summit marks a critical step in ensuring that AI technology is harnessed for the benefit of society while mitigating potential risks.
