Some thoughts on the current UK policy focus on AI

I was recently asked by Research Fortnight to comment on the UK government's recent policy focus on AI. I gave some quick thoughts and inevitably, as usual, not all of them ended up in the article, so I'm using this opportunity to share my full comment in the hope it is of interest. I write as a research and innovation policy scholar with no deep expertise in the social and economic impacts of AI but with some insight into the broader dynamics of emerging technologies and government policies.

AI as a research area has been around for many decades, but the recent furore about generative AI models like ChatGPT means that it has reached a sort of tipping point in public attention. And if something is getting a lot of public attention then it is likely to also be on the minds of politicians. So that’s the first possible dynamic to consider here.

But there’s also hype cycle aspect to this. All emerging technologies are subject to hype by their promoters, and all governments are sensitive to such hype. In particular, recent UK governments have appeared to be very susceptible to Silicon Valley tech boosterism, with several commentators pointing out how much more political attention is paid to Silicon Valley leaders than the leaders of much larger, more economically and socially significant mainstream sectors. Again, this is by no means just a UK phenomenon - but because of the cultural cringe most UK politicians and policy wonks on both the right and left have towards the US, it can seem particularly strong here.

Finally, and not least, there’s the more hard-headed policy interest in AI – first, in not being left out in some kind of 'global race' for mastery (especially in this new era of great power competition); and second, the realisation that technologies under the broad heading of AI are coming into use and that this creates policy and regulatory challenges.

Some of the understanding here seems from the outside to be a bit muddled (the global race, if that is even the right metaphor, is by no means a zero sum race; and what are we seeking mastery of, the underlying technologies or key economically or socially impactful applications?). But this is absolutely the role of government - and the UK does have great research strengths in questions of responsible innovation that could and should be applied to AI.

There's a danger that hype or the political need to be seen to be doing something will drive the development of superficial or sub-critical activities. But it is right to be paying significant policy attention to AI and what I’d like to see is some very transparent governance about how priorities are being set around AI research, innovation and regulation, with explicit discussion of the tradeoffs between different objectives, and with really meaningful public engagement built in. In this regard it's great to see UKRI today announce a ￡31m investment in a UK consortium that will hopefully do some of this.

Finally, there's one more political angle. AI is of course also a sexy topic ahead of the election and one that, unlike most other areas of economic policy, does potentially allow the government a priority theme that helps them to swerve unwanted debates about the consequences of Brexit.

??