Artificial intelligence (AI) is heralded as the next general-purpose technology. A bold claim that puts it in the same bracket as the internet and electricity.
You can understand why, as AI technologies aim to reproduce or surpass abilities (in computational systems) that would require ‘intelligence’ if humans were to perform them.
AI is already transforming the global economy, drawing on vast datasets to perform complex tasks such as allowing people to communicate across the globe using speech recognition and translation software.
You and I come across AI most regularly through the recommendations for products and films we see as we browse the web or stream services.
AI is also transforming health services in the prediction, prevention and treatment of diseases. It’s commonplace in applications such as financial trading, environmental monitoring and driver-assistance.
With so many applications, you can see why AI has so much potential. Many believe countries leading in AI could gain an extra 20 to 25% of gross domestic product over the next decade.
We haven’t realised the benefits of AI
However, the profound impact of AI on the economy hasn’t yet been realised and will rely on further research, development, innovation and commercialisation.
I also see that the growing use of AI raises genuine concerns about increased surveillance, manipulation of news feeds, automated decisions being made about us that are based on biased information and even machines replacing people’s jobs.
A wide range of abilities are required by ‘computational systems’ for them to be intelligent, such as reasoning, planning, learning, adaptation, sensory understanding. No single AI technology can do all of this.
Even when several AI technologies are combined, it is not always possible to achieve all the abilities needed for AI to work reliably in the way we want and need it to.
Much more research and innovation are required to realise the promise of this technology to the economy and society, and this is where UK Research and Innovation (UKRI) comes in.
Working across disciplines to develop AI
The growing importance and reach of AI prompted us to review our work in this area, but not just in one council. We looked at AI across all councils, how it featured in our portfolios, how our communities approached it, and what the opportunities and gaps were.
UKRI councils came together to undertake what I believe was the first subject-based cross-UKRI review, publishing our findings in ‘Transforming our world with AI’.
Our findings have been a major influence on the approach we are now taking to AI across UKRI.
Advances come from applying AI to specific questions
AI is not one technology but a collection of technologies with a range of properties and abilities that currently can’t readily be transferred across sectors or applications.
The nature of data-dependant technologies such as AI means that innovations and improvements to the technology often come through it being applied to a specific problem or application.
One of the most significant findings of our review was that AI is advanced by tackling real world problems in a responsible manner. It was apparent that, for this technology to flourish, we must bring together computer science, mathematics, data, humanities and social science expertise with innovators and application-based expertise for potential uses across all the areas supported by UKRI. For example, AI that can work with healthcare, environment, lifestyle and other datasets collectively to predict resilience to health conditions.
AI is an area that will truly benefit from this one-UKRI approach.
Communities need to view AI as multidisciplinary
Just as we are thinking differently about AI, our communities also need to think differently about AI.
They must see it as a multidisciplinary endeavour if it’s going to have a positive transformational impact on our society and economy. We need our communities to come together to research, develop and innovate in AI in a responsible manner, developing more advanced AI technologies that are trusted and work reliably in a way that is acceptable to the general public.
The additional funding we secured from the Treasury for centres for doctoral training in AI is the most important of the opportunities we have developed through working across UKRI.
AI skills are in huge demand across the world, business and academia both struggle to recruit and retain top AI talent.
This skills shortage is highlighted in the National AI Strategy, which sets out how we must attract and retain the best and most diverse people to build AI, use AI and be inspired by AI.
Centres aim to fill skills gap
We built a strong case to demonstrate the role of centres for doctoral training in filling this skills gap and setting the UK on a path to remain a science and AI superpower. By doing so we have secured £117 million from the Treasury for a new tranche of AI centres for doctoral training that will train cohorts of graduates in the application and implications of AI technologies, providing opportunities for peer-to-peer learning and for students to work directly with industry on AI challenges.
This was a fantastic opportunity not just to grow UKRI investment in AI but for UKRI to demonstrate the value of multidisciplinary approaches to AI and of growing numbers of people with AI skills across our communities.
Those leading on AI across our councils worked collaboratively to shape this investment, building an opportunity that meets the needs of our individual communities and embeds multidisciplinary working in the training experience of these future AI leaders. I’m immensely proud of the collaborative approach taken to developing this funding opportunity. If you would like to find out more about the work we’ve been doing in AI and the direction we want to take, read ‘Transforming our world with AI’.
Top image: Credit: UKRI
Kathryn Magnay is currently a Deputy Director of Cross-Council programmes at the Engineering and Physical Sciences Research Council (EPSRC) with responsibility for EPSRC’s ‘AI, Data and Digitalisation: Driving Value and Security’ priority.
In this role Kathryn has led the UK Research and Innovation wide review of artificial intelligence and has oversight of EPSRC’s ‘AI, Robotics, and Digital Security and Resilience’ portfolios.