UCL has launched an innovative strategy detailing how Artificial Intelligence (AI) should be used as a transformative technology to make a positive impact on the planet.
AI for People and Planet was launched by Professor Geraint Rees – UCL’s Pro-Vice-Provost of AI – at the Science Museum in London on Tuesday 10th September. It encapsulates the belief that the purpose of research and innovation in the sector is ultimately to benefit people and societies around the world and to have a positive impact on the planet.
The launch event included a panel discussion around issues relating to AI in healthcare, society, ethics, economics and the current state of AI. As well as UCL experts, the panel also included Professor Thore Graepel, Research Group Lead of Google DeepMind and Azeem Azhar, board member of the Ada Lovelace Foundation, venture partner at Kindred Capital and advisor to Fabric Ventures. It was moderated by Libby Kinsey, co-founder of UK AI ecosystem promoter Project Juno.
Additionally, to support the strategy, the UCL Centre for Artificial Intelligence is opening in Holborn, central London, which will be part of UCL’s Computer Science department and headed up by Professor David Barber, Director of the Centre and Professor of Machine Learning.
The centre will house over 100 researchers carrying out foundational research into AI, supporting the creation of spinout companies driven by deep technical advances and amplifying UCL’s impact in academia and industry.
AI is widely used in sectors ranging from health and education to science and art, with consumers regularly using machine learning without realising. Everyday tasks such as applying for bank loans and running online searches are governed by AI, which also serves tailored online adverts based on past activity.
The potential for AI technologies is vast, with upcoming challenges which will need addressing including how this can affect industries such as transport through autonomous vehicles, agriculture through automated farming and healthcare through personalised medicine. It also has the potential to change the way we work through supporting and enhancing productivity.
Concerns about the use of AI technology range from using it to create fake videos, known as deepfakes, to increased energy requirements as more complex machinery is developed. AI has a vital role to play in the solution, through monitoring the environment and reducing dependence on fossil fuels.
“The question we need to address is how to sustainably manage the global growth of AI,” commented Professor Rees. “We need to be thinking about the interplay between AI systems and the planet inhabited by humans and other animals, and how they coexist.
“The human dimension is key; AI should be about remedying human actions and bias rather than replicating them.”
The AI for People and Planet strategy provides a solid framework for work by UCL and its partners within the context of the Industrial Strategy Grand Challenge on AI and Data. Its vision is framed around five key points: the human dimension, inclusion and diversity, interdisciplinarity, open science and misuse.
The strategy covers foundational AI – which sits at its core through the new AI Centre – the application of AI and responsible AI for urbanisation and society. Reflecting the interdisciplinary nature of UCL, the strategy draws on expertise from departments and faculties across UCL including (but not limited to) the Gatsby Computational Neuroscience Unit, the UCL Department of Science and Technological Studies and the Institute of Education (IoE).
Professor Barber continued: “AI is already making a big impact in our daily lives and has the potential to help us address some of the greatest challenges our societies and the environment face.
“In the future AI can replace manual labour and free people to be more creative, more caring and more concerned. However, to realise this potential we need to continue to innovate in fundamental research. AI is still very much in its infancy and the UK needs to continue to play an important role in developing next generation AI technologies.”
Professor Rees added: “Developing AI sustainably also means looking at carbon emissions and power consumption of the industry. Experts in the field need to be looking at how algorithms can be used in a positive way to lower these, particularly given that we are creating machines and processes which require large amounts of power to operate.”
The UCL Centre for Artificial Intelligence will be at the forefront of global research into AI, with its academics creating new algorithms to support science, innovation and society. It will be key in ensuring that the UK remains at the heart of AI innovation and supports productivity growth.
UCL is central to UK efforts to train next generation AI specialists, with two major government-supported PhD training programmes in AI for Healthcare and Foundational AI launching this academic year, totalling over £20m in government and industry investment. Additional support from CISCO, Google DeepMind and Accenture has enabled the creation of fully funded places on Masters programmes in AI related subjects for students from under-represented backgrounds.
This communication was first published 11 September 2019 by University College London