UKRI AI Centre for Doctoral Training in Foundational Artificial Intelligence
Lead Research Organisation:
University College London
Department Name: Computer Science
Abstract
The UK has had remarkable success in producing new algorithms in Artificial Intelligence. This has driven two processes -- one is the adoption of existing large-scale data analysis tools to industry problems. The other is the acquisition by large tech companies of UK talent, with the intention that it will significantly shape the future commercial landscape.
Whilst there has been dramatic recent progress in AI, the systems are still far from being universally useful. The tech giants are well aware of this and are investing heavily in teams and algorithms that can address these fundamental AI challenges.
It is vital that the UK retains the ability to create new AI technologies and does not become merely a consumer and user of AI technologies produced by others. This training centre will spearhead this effort by addressing the core research problems and training students to become the next generation scientists and entrepreneurs.
Whilst there has been dramatic recent progress in AI, the systems are still far from being universally useful. The tech giants are well aware of this and are investing heavily in teams and algorithms that can address these fundamental AI challenges.
It is vital that the UK retains the ability to create new AI technologies and does not become merely a consumer and user of AI technologies produced by others. This training centre will spearhead this effort by addressing the core research problems and training students to become the next generation scientists and entrepreneurs.
Planned Impact
WHO MIGHT BENEFIT?
Science : A key beneficiary is the science community itself. The challenge of advancing AI to incorporate more knowledge about the human world, the objects therein and to thus be able to interact with us in more natural ways, is difficult but exciting and potentially hugely rewarding. As an example, a machine that can read and comprehend texts and form a unified representation and understanding of the information within those scattered texts has the potential to transform areas such as medicine as otherwise unconnected ideas can be brought together to form new insights. Making new AI technologies that can explain how they take their decisions and rationally justify them is of great importance is certain sectors such as medicine.
Society : Making explainable and transparent AI is clearly of great concern to the general public. This is a relatively new area of research and the CDT is well placed to make progress on this, particularly in partnership with our partner the Alan Turing Institute. DeepMind have a research programme that is attempting to address this and we expect to benefit from interactions with them as well as other researchers.
Startup Companies : We expect a number of AI startups to emerge from the CDT. This will directly benefit the UK economy by providing employment and investment in the UK. The UCL Computer Science Department has a good track record of making AI startups, with the management of the CDT being themselves successful AI entrepreneurs. There is significant infrastructure in place to help train and support PhD students to generate such companies, with a wealth of local expertise from academics and other students that have successfully achieved this.
Companies : CDT graduates are expected to be leaders in new AI technologies, going well beyond the application of existing data analysis tools. For example, we expect CDT graduates to provide their future employers will the ability to design new products based on the highly specialised training the students will have received. Through our CDT industry placement scheme, companies will also benefit from having students will world-leading AI talent to help them drive innovation.
Government and Public Policy : UCL runs a public policy placement scheme with UK government departments. This provides an opportunity for students to contribute to policy and decision making of potentially national significance.
Medicine : We expect to make advances in healthcare which will make more efficient use of hospital resources. We will also make advanced algorithms that will improve medical diagnosis and eventual treatment of disease.
HOW MIGHT THEY BENEFIT?
Our industry advisory board will ensure that research and teaching remain relevant.
The AI Centre at UCL will regularly host events and invite industry to view the outputs of research from the Centre. The CDT will be embedded within this larger AI Centre and students will therefore have the opportunity to engage with and get expose to industry.
Through our annual AI pitch day, investors will have the opportunity to fund student ideas. UCL has a unit dedicated to supporting AI spinouts, including seed funding, mentoring and space.
The CDT aims to place all research and results in the public domain. In addition to academic publication channels (the major AI conferences and journals) we will reach out to more popular science and media sources (our UCL advisor Jack Stilgoe is a Guardian columnist on ethical AI).
It is standard in AI research to produce public software alongside publications. To maximise impact we will mandate (wherever possible) that software is openly available.
The CDT will maintain a high profile website and a manage a corresponding social media profile. We have a dedicated media officer in the Computer Science Department at UCL who will deliver this.
Science : A key beneficiary is the science community itself. The challenge of advancing AI to incorporate more knowledge about the human world, the objects therein and to thus be able to interact with us in more natural ways, is difficult but exciting and potentially hugely rewarding. As an example, a machine that can read and comprehend texts and form a unified representation and understanding of the information within those scattered texts has the potential to transform areas such as medicine as otherwise unconnected ideas can be brought together to form new insights. Making new AI technologies that can explain how they take their decisions and rationally justify them is of great importance is certain sectors such as medicine.
Society : Making explainable and transparent AI is clearly of great concern to the general public. This is a relatively new area of research and the CDT is well placed to make progress on this, particularly in partnership with our partner the Alan Turing Institute. DeepMind have a research programme that is attempting to address this and we expect to benefit from interactions with them as well as other researchers.
Startup Companies : We expect a number of AI startups to emerge from the CDT. This will directly benefit the UK economy by providing employment and investment in the UK. The UCL Computer Science Department has a good track record of making AI startups, with the management of the CDT being themselves successful AI entrepreneurs. There is significant infrastructure in place to help train and support PhD students to generate such companies, with a wealth of local expertise from academics and other students that have successfully achieved this.
Companies : CDT graduates are expected to be leaders in new AI technologies, going well beyond the application of existing data analysis tools. For example, we expect CDT graduates to provide their future employers will the ability to design new products based on the highly specialised training the students will have received. Through our CDT industry placement scheme, companies will also benefit from having students will world-leading AI talent to help them drive innovation.
Government and Public Policy : UCL runs a public policy placement scheme with UK government departments. This provides an opportunity for students to contribute to policy and decision making of potentially national significance.
Medicine : We expect to make advances in healthcare which will make more efficient use of hospital resources. We will also make advanced algorithms that will improve medical diagnosis and eventual treatment of disease.
HOW MIGHT THEY BENEFIT?
Our industry advisory board will ensure that research and teaching remain relevant.
The AI Centre at UCL will regularly host events and invite industry to view the outputs of research from the Centre. The CDT will be embedded within this larger AI Centre and students will therefore have the opportunity to engage with and get expose to industry.
Through our annual AI pitch day, investors will have the opportunity to fund student ideas. UCL has a unit dedicated to supporting AI spinouts, including seed funding, mentoring and space.
The CDT aims to place all research and results in the public domain. In addition to academic publication channels (the major AI conferences and journals) we will reach out to more popular science and media sources (our UCL advisor Jack Stilgoe is a Guardian columnist on ethical AI).
It is standard in AI research to produce public software alongside publications. To maximise impact we will mandate (wherever possible) that software is openly available.
The CDT will maintain a high profile website and a manage a corresponding social media profile. We have a dedicated media officer in the Computer Science Department at UCL who will deliver this.
Organisations
- University College London, United Kingdom (Lead Research Organisation)
- Huawei Technologies (UK) Co. Ltd (Project Partner)
- Dynium Robot (Project Partner)
- nVIDIA, United States (Project Partner)
- Digital Surgery (Project Partner)
- EntrepreneurFirst (Project Partner)
- ASI Data Science (Adv Skills Initiative) (Project Partner)
- SCM Advisors (Project Partner)
- Capital Enterprise LLP (Project Partner)
- Adobe Systems Incorporated, United States (Project Partner)
- Connected Digital Economy Catapult, London, United Kingdom (Project Partner)
- Julia Computing (Project Partner)
- DeepMind (Project Partner)
- BenevolentAI (Project Partner)
- Cisco Systems UK, United Kingdom (Project Partner)
- Albion Capital (Project Partner)
- Microsoft Research Ltd, United Kingdom (Project Partner)
- Vodafone Group Services Ltd, United Kingdom (Project Partner)
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
EP/S021566/1 | 31/03/2019 | 29/09/2027 | |||
2267930 | Studentship | EP/S021566/1 | 22/09/2019 | 21/09/2023 | Samuel Cohen |
2251002 | Studentship | EP/S021566/1 | 22/09/2019 | 21/09/2023 | Jas Semrl |
2251012 | Studentship | EP/S021566/1 | 22/09/2019 | 21/09/2023 | Yihong Chen |
2250981 | Studentship | EP/S021566/1 | 22/09/2019 | 21/09/2024 | Dan Stoddart |
2251009 | Studentship | EP/S021566/1 | 22/09/2019 | 21/09/2023 | Felix Richard Biggs |
2251565 | Studentship | EP/S021566/1 | 22/09/2019 | 21/09/2023 | Jingwen Wang |
2250955 | Studentship | EP/S021566/1 | 22/09/2019 | 23/12/2023 | Augustine Mavor-Parker |
2253986 | Studentship | EP/S021566/1 | 22/09/2019 | 31/12/2023 | Luca Morreale |
2251578 | Studentship | EP/S021566/1 | 22/09/2019 | 21/09/2023 | Changmin Yu |
2252836 | Studentship | EP/S021566/1 | 22/09/2019 | 21/09/2023 | Jakob Zeitler |
2408233 | Studentship | EP/S021566/1 | 26/09/2020 | 29/09/2024 | Denis Hadjivelichkov |
2408383 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Oscar Thomas Key |
2409434 | Studentship | EP/S021566/1 | 27/09/2020 | 31/12/2024 | Oliver Slumbers |
2408098 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Reuben Adams |
2408309 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Jean Heidar Kaddour |
2408394 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Robert Kirk |
2409398 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Antonin Florentin Schrab |
2409446 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Sicelukwanda Njabuliso Zwane |
2408220 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Yue Feng |
2409392 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Yicheng Luo |
2408302 | Studentship | EP/S021566/1 | 27/09/2020 | 29/09/2024 | Alex Hawkins-Hooker |
2419792 | Studentship | EP/S021566/1 | 30/09/2020 | 29/09/2024 | Mirgahney Husham Mohamed |
2425997 | Studentship | EP/S021566/1 | 30/09/2020 | 29/09/2024 | Linqing Liu |
2409437 | Studentship | EP/S021566/1 | 30/09/2020 | 29/09/2024 | Yuchen Zhu |
2570923 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Emilio Dara McAllister Fognini |
2570649 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Laura Eline Ruis |
2570828 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Zak Stefan Morgan |
2570927 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Harry Jake Cunningham |
2570758 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Harry Songhurst |
2570916 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Aengus Lynch |
2574460 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Nikolay Dagaev |
2584610 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Laura Eline Ruis |
2570783 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Akbir Khan |
2570920 | Studentship | EP/S021566/1 | 30/09/2021 | 29/09/2025 | Rokas Bendikas |
2584277 | Studentship | EP/S021566/1 | 01/11/2021 | 31/10/2025 | Laura Eline Ruis |
2575973 | Studentship | EP/S021566/1 | 01/11/2021 | 31/10/2025 | Mariia Naslidnyk |