2023–24年度报告

2023–24 Contents /05Page 36 - 432024-25 planning /01SummaryPage 04 - 09 /03Page 16 - 29Progress againststrategic objectives Page 10 - 15Introduction /04Page 30 - 35Looking Ahead Page 44 - 47Annex 1: Steering groupmembership Summary ai@cam is theUniversity ofCambridge’s mission tocreate AI innovationsthat serve science,citizens, and society.“ AI innovations that servescience, citizens, and society. Breadth and diversity ofengagement in the AI-deasscheme: 100+ Number of people consulted tocreate Cambridge’s AI vision 7373 applications from27 Departments 2022’s ai@cam review identified five core functions neededto deliver Cambridge’s vision of AI that serves science,citizens, and society: a dynamic research community;partnerships that centre societal needs; spaces that helpprojects to spin up and respond to areas of need; wideraccess to AI knowledge and skills; and connections to local,national, and international priorities for AI. These areas ofneed have been central to ai@cam’s first year of activities: Collaborations with University of Cambridge Developmentand Alumni Relations (CUDAR) and the StrategicPartnerships Office are building connections withstakeholders in industry and philanthropy to help scaleour activities and our next year of operations will bringfurther work in this area. 4646 research projects from20 Departments across6 Schools onboarded to theDawn system. £750kResulting in5 fundedchallenges, with a totalvalue of £750k ANNUAL REPORT 2024 • AI@CAM→Our AI-deas initiative launched five ambitiousresearch programmes that bring togetherresearchers from 19 Departments across all 6Schools using AI to to tackle critical challenges inmental health, fertility, public services, languageequity, and nature and climate.ANNUAL REPORT 2024 • AI@CAM →Our Policy Lab has catalysed collaborations acrossfive University research centres, connecting theUniversity’s expertise in AI, policy, and publicdialogue to the challenges faced by policymakerslocally, in Westminster and Whitehall, andinternationally. 19 400+Number of participants inworkshops and events Working across 19Departments spreadover all 6 Schools. →Responding to an opportunity presented byResearch Computing Services’s new HighPerformance Computing (HPC) system, our HPCPioneer Project scheme has helped 46 researchprojects from 20 Departments get off the groundwith access to the University’s world-leadingcompute resource and support from dedicatedmachine learning engineers. Our firstyear ofoperations hasestablishedai@cam as anincubator forinterdisciplinaryAI that addressesreal-world needs.“ £190k+ External incomegenerated →ai@cam has also played an enabling role,supporting the AI community to respond to externalopportunities, such as UKRI funding schemes,providing support for large-scale funding bids, anddeveloping philanthropic and industry partnerships. These initiatives are rooted in both community need andpublic interest. They are shaped by over 400 engagementswith researchers, policymakers, partner organisations,and the public across 2023-24, and have been guided byour cross-institutional Steering Group. £100M+Applications pendingor in development. SECTIONEngagement with early career researchers has been astrong feature of our first year. Out of 73 expressionsof interest in our AI-deas call for projects, 41 came fromearly career staff, and 26 (of 46) of our HPC pioneersare PhD students. Feedback from these engagementssuggests that our approach to convening the communityhas helped researchers find new collaborators and unlockaccess to resources.SECTION ai@cam’s vision for aworld-class AI researchresource ai@cam offers a mechanism for the Universityto be a global force in AI research, a world-leader in education, and an engine forinnovation that delivers social benefit. Byinvesting in core research capabilities andcreating collaborations that drive progressin areas of scientific and social concern, theUniversity can build a world-class AI researchresource. By building these pillars, Cambridge canensure it has the talent, connections, andcollaborative culture to scale its research andeducation activities, leading a new wave of AIinnovation. This will continue to be a priorityfor the ai@cam mission in the coming years. This vision is built on three pillars of activity: →Growing our research capability:cluster hiring of new faculty, machinelearning engineers, and PhD studentscan create a critical mass of expertisein AI technologies and their safe andeffective use. →Building a core of activity around world-class compute: new HPC systems offeran opportunity to convene core AI teamsaround world-leading research facilities. →Creating connections for impact:interdisciplinary engagements with policy,industry, civil society, and the public offeropportunities for creating new challenge-led research programmes and enhancingthe impact of existing activities. IntroductionSection2 Supporting the next generationof AI innovation