The UK is a world-leader in the development of artificial intelligence technology. Our record is such that, even when our home-grown companies have been bought by foreign investors, they have often retained the staff and presence right here in this country – for example the AI company, Deepmind, was purchased by Google, but retains its headquarters in King’s Cross. But our recent review for the Government found that this is a situation far from guaranteed to continue unless Ministers and industry act. If they do, it could have considerable benefits for the economy, including in the operation and maintenance of our infrastructure.
It’s easy to see why this is so important: worldwide, the market for AI solutions could be worth more than £30billion by 2024. We’re already well-placed to make the most of this – a Coadec report recently found that a new AI start-up was founded almost on a weekly basis over the past 36 months. But that same study also demonstrated how the rest of the world is catching up, with only one in 10 UK AI firms in the late ‘growth’ capital stage of their development compared to one in five in the US.
So how do we keep up with this ever-changing landscape and compete on the global stage? The recommendations in our review covered four main themes:
Access to data is essential to cultivating innovation and new applications for AI technology, particularly among small start-up companies. But clearly that needs to be done in such a way that keeps personal records secure, and ensures the public can have confidence that their information is safe.
We’ve recommended the creation of “data trusts”, legal frameworks through which data can be shared in a mutually beneficial, and secure, way. We recommend that the development of such data sharing relationships be supported by a new Data Trusts Support Organisation that will initially be supported by government funding and hosted by an appropriate independent organisation such as the Royal Society or the Open Data Institute.
This wouldn’t happen overnight – in fact, these relationships would have to be built up sector-by-sector to test the idea and see how it would work in practice. The transport sector is a strong candidate to be one of the first for this: many organisations could benefit from sharing data – just look at how open data around London’s transport network has led to such apps as CityMapper for those looking to commute across the Capital – but where the practical processes have not yet been trialled. The health sector is another, although of course that is even more fraught with the really difficult problems of sharing data that is highly confidential and private: but if we can crack the problem in a meaningful way it could bring huge benefits to both society and the UK economy
If we’re to retain our place as one of the leaders in AI technology, we need to secure the supply of skills and invest in the next generation of pioneers, encouraging more people from a more diverse range of backgrounds to consider it as a career. Our review has recommended an industry-funded Masters programme in AI, 200 new places for PhDs at UK universities, and market research to develop conversion courses that meet employers’ needs. Already companies involved in infrastructure including Siemens, Shell, Tata Steel and National Grid are looking to get involved in the latest thinking in this area through the Alan Turing Institute’s regular data study groups: this could provide them with another means of being directly involved in the next AI innovations. And if this is going to work, we need students and staff from a range of backgrounds and with a range of skills, to ensure the data we use and the AI we develop isn’t instilled with the natural biases of those involved in designing it.
Already the UK has an enviable record in the research of this new technology – but more needs to be done to maximise AI research and commercialisation. That’s why we’ve recommended that the Alan Turing Institute become our national institute for artificial intelligence and data science, and that universities promote standardisation in the transfer of Intellectual Property to enable that to happen. But we’ve also urged collaboration across the UK research community to negotiate for computing capacity for AI research, helping to reduce its cost for everyone.
And finally, we need to further support the uptake of AI. Already the Government’s Industrial Strategy Challenge Fund helps bring the technology to areas where it wouldn’t otherwise be commercially viable to do so. So far, that’s included mapping capacity and logistics in the social care sector to improve provision, and improving the resilience of UK industry to cyber-attacks.
But during our review, we found that this fund could also apply to the use of AI technology to create integrated transport for smart cities, and improving the efficiency of our infrastructure. With that in mind, I’m looking forward to seeing the findings of the National Infrastructure Commission’s New Technology study, which examines how digital technologies can be applied to infrastructure. This will include looking at such opportunities as reducing leakages in our water supply, and smart systems for managing traffic flows.
We’ve only scratched the surface of what AI can do for us, for our infrastructure, and for the economy at large. The pace of innovation is already fast – and it’s only going to get faster. We are already in a strong position to retain our place as one of the leading countries for this technology, but our review shows that there’s more we can do, and there are plenty of opportunities for players across the infrastructure sector to get involved. That’s why we are recommending the establishment of an AI Council to bring together all the stakeholders in this revolutionary technology to ensure the recommendations in our report are taken forward for the common good.
Professor Dame Wendy Hall is co-author of “Growing the artificial intelligence industry in the UK“, an independent review for Government