Kanishka Narayan MP: Minister for AI and Online Safety, on how the UK can win in AI!

0:00Kanishka Narayan:Hello, welcome back to Scaling Your Retro. I'm here with Kanishka Narayan, Minister for AI and Online Safety, which is a very interesting place to be at the moment. There's been lots of exciting announcements about AI. And of course, AI is just the most exciting place to be, especially from my perspective. Today, there's a really exciting announcement. It's a big plan to upskill 10 million people in AI skills. I think it's the most ambitious, audacious training program that has kind of embarked on since it is AI Bulsome. why is now the time to be embarking this huge training plan? Sam, firstly, can I just say it's so exciting to be chatting.

0:38I've been following you on socials and you're doing an amazing job at shifting where I think content and culture is in Britain when it comes to technology as well. And so the announcement today is huge because from my point of view, AI is a transformative economic opportunity for countries right across the world. And the question is, does Britain get on the bus and capture that opportunity or not. What we're trying to do today is to say that we've got to capture that. And by doing so, we've got to make sure that, frankly, unlike the last wave of SaaS and software, we need to make sure that the opportunity of AI is spread right across the country.

1:09That's why we want to reach 10 million people. We were on target for seven and a half. We're looking like we're going to burst through that target. So we've now said we're going to set our sights even higher. 10 million people across the economy in the NHS and big businesses and small and medium-sized companies trained out for the future of AI.

1:25Kanishka Narayan:And what can this unrop, do you think, if it goes whether you kind of succeed in your mission? Yeah, I think there's a simple thing, which is that economic growth of productivity. So real wages for people right across the country being higher because they're able to use AI to augment themselves. But the big opportunity for me is not just economic, it's actually substantively about communities right across the country. So for a kid growing up in, whether it's South Wales, close to my heart or right across the UK, to be able to scale up in AI and as a result access jobs that they otherwise couldn't in their community, that will be an absolutely fantastic outcome.

1:58Kanishka Narayan:And do you worry about the, I guess, the gap widening between some of the classes in society as we see the skills in AI becoming more and more in demand and they are becoming more and more powerful for those who are equipped with them? I guess, was that part of the rationale that some people are at risk of being left behind? I think a huge part of what we're trying to do is to say that the entire AI opportunity only works if it's spread right across the country and we buy into it as a country. And so a lot of what today is about is to say that we have agency on this question. It is not necessarily the case that AI will increase inequality.

2:32It is not necessarily the case that AI will spread opportunity. And so we are exercising agency by making sure people are skilled up right across our economy. And by making sure that jobs which we think are going to be lasting long-term jobs, which use human labor and manual dexterity, our health care support, for example. The NHS has signed up to be one of the employers supporting us with this mission as well. So that's the critical thing. We want to make sure we're supporting Google and Jobs of the Future in every community to make sure that the gap between those who are best able to use it and those who are supported is as narrow as possible.

3:02Kanishka Narayan:And when we talk about AI skills, it's quite a broad term, I guess. What kind of skills are you imagining that people will be actually upskilled in? There's the sort of first rung of the ladder. You know, I did a course on mastering prompting as well. All of us can kind of do with it a little bit as partly a thing you learn from practice. But here there's kind of the building blocks of it, which is if you're coming completely fresh to it, you can get on the runs of the ladder and feel more confident in using it in your day-to-day life and day-to-day work. All the way through to the fact that organizations, we've got Accenture, we've got Multiverse, we've got a range of large enterprises saying, what are the ways in which our workforce can use it to integrate AI into their existing workflows?

3:42That's really where the magic is. It's not just in sort of standalone tools. it's about the glue that ties different tools together different models together so it fits into people's day-to-day work

3:52Kanishka Narayan:and are there any AI tools that you've experimented with that you really like or any AI tools that have become part of your day life? Gemini I feel like you know as with a lot of people Chagicutee and then Claude were kind of great I used to love the tone and character of Claude but now I increasingly feel Gemini does the job really well um i sometimes uh do a bit of uh content editing in generation nowhere near as good as you uh but with the team you know we sometimes use uh okus clip and uh yes and similar tools uh and then i very briefly dabbled with uh cursor as well uh just for kind of low-level coding nice do you feel like you have a responsibility to try and test these tools or do you just do it because you're interested and you love it yeah i think i think i've uh i i personally feel kind excited about trying to get a sense of how I can do a better job day to day.

4:45And so part of it is personal intrigue. Yeah, part of it is, as you say, making sure that in the role that I'm doing, especially given it's both AI and online safety, that I'm focused on what the opportunities are from these tools for the country, but also where we want to make sure that we're working responsibly to make it a an opportunity that people trust.

5:05Kanishka Narayan:And how do you think the UK UK tech is doing in this sort of ai race hey the big thing is we're firing on all cylinders it's absolutely astonishing i think last year was a crack in the air for fundraising i think this quarter is going to beat all the records uh already you just um i know you covered the synthesis announcement recently as well uh that is part of a wider thing but if i just look across all the big inputs you look at capital you record capital fundraising we've ramped up pvb uh in terms of what the government is doing there as well if you're looking uh compute and the four and a half five months i've been in this job we've announced 56 billion of private investment in data center infrastructure and computed this country if you look at talent not only are we building this great skills and pipeline in the uk but we are now a country that says whenever whatever others are doing we are open to the best talent in the world coming here and building great companies creating jobs we have created a specific ai stream in the global talent visa program reimbursing fees making it a really smooth experience for folks and similarly on on data We've got some great public initiatives on making sure that in terms of public research into health care in particular, we're right at the frontier.

6:09So whatever all those inputs, they're all working. But the big thing from my point of view is I'm starting to feel a sort of cultural bike shift where all of that is now coming together. The density of that means that we are talking about this with a sense of energy and enthusiasm that I think is really foundational.

6:25Kanishka Narayan:Yeah, I think the optimism is strange. It is in a lot of narrative about how far behind the UK is or how far the Europeans and how we're not coming up to capture. but it feels like last this modeling has changed we're seeing these category defining businesses come out runden uh which i think is amazing yeah so there's so much that's going right so much is going it's moving forward yeah what do you think we need to do more if we'll be better at in the uk the the single big thing that i try to do in my job and and would love the ecosystem as a whole is to dream a sort of order or two of magnitude higher yeah let's more be happy trying to say that hey we're the third best in the world between the US and China and the rest I want to be shooting for the top and I know the government as a whole does and we need this relentless mindset every day that all we're trying to imagine here for the country is a really wonderful, fantastic, large-scale vision for what UK AI can do in the future not just in terms of the global scale but in terms of the domestic spread in every part of the country.

7:22Kanishka Narayan:And we were seeing the UK government take quite a practical approach to partner with large tech organisations I think there was a partnership announced with OpenAI. There was a memorandum of understanding signed with Anthropic last year. We're now seeing that partnership widen and that Anthropic is being used to, it's going to roll out like an AI assistant specifically in the first use case for employment cases. Can you talk a bit about that partnership with Anthropic? Why do you think it's important? Yeah, the fundamental focus for the government is to make sure that those who have not historically benefited from frontier technology get to do so.

7:52So the Anthropic Partnership for an AI assistant on gov.uk is really about saying that historically, the experience of public services has not been the most pleasant experience for a lot of people. When someone is in a challenging place trying to look for a new job and desperately seeking out lots and lots of forms and a not great UK government website, what we've done is, of course, we've made huge strides in improving that website. And now we want to make sure that experience is even better so that we are putting Frontier Technology, the service of dignity of people right across this country.

8:20and then in terms of partnerships more generally what we're trying to do is to say that there are some large companies the the foundation models are going to be really critical in terms of the overall stack so we must partner with those who are really the best uh the best in the in that business but alongside that we want to back uk small tech and that's why the initiatives on sovereign ai on an advanced market remember for novel shapes on a huge amount we're doing not least this week in defense we've announced the 20 million startup fund focus on uh procuring from the best British startups as well.

8:51And so we've got this balanced economy we're trying to build where we are right at the frontier and in some cases with the large companies that are brilliant at doing what they're doing. But alongside that, we're really building this fantastic ecosystem of British jobs.

9:03Kanishka Narayan:And I think that part of you is really interesting. One of the things that we're seeing at the moment, especially in the last week, France, we're now seeing the shift towards away from international and American tech companies. France is now replacing them with homegrown French or European ones. Given the political landscape at the moment, Do we think we as a country need to be rethinking some of the partnerships that we have, specifically with American big tech and trying to prioritize either homegrown solutions or at least European ones? I guess I approach this largely with a sort of abundance mindset, which is that you can either try and say, I'm going to pick this or that.

9:35We can say, I expect the overall scale of demand to be so high that the one part of the building blocks of how we make sure we have enough of this stuff is, yes, to just partner with those who have capacity and scale. and we have great partners in the United States that we're working with to make sure that we're delivering of opportunity jobs and great experiences for people in the UK. But alongside that, we want to make sure the British startups capture a bunch of the opportunity as well. And so I don't see a huge trade-off in these things. The reality of it is we need to make sure that AI is supporting our experiences in lots of public services and it's supporting our economy in the private sector.

10:11We can do that with our partners in the United States. We can do it independently and we can do that, both of those things,

10:17Kanishka Narayan:together amazing and what advice would you give to somebody you know maybe in the south of wales or anywhere around the uk who's looking to upscore themselves in a who's maybe feeling a bit left behind feeling worried about being left behind because the advance of ai the new still says demand that is being required yeah what advice would you give them to get started into the journey of ai yeah i think so uh one of the big announcements today is that you can go on the airstillshub.org.uk we have a bunch of free training courses for people right across the country. So the biggest piece of advice in any context is to get started.

10:50So you can start today. But the second thing is, this is truly something that when we talk about the technology, we forget that this is something that each of us can use in our daily lives. When we use Google Maps, we call it Google Maps. We don't call it AI that has ended up helping us navigate cases. When we use a spell check, we don't talk about natural language processing. We talk about checking the spelling of what we're writing and so how we can make sure that we're thinking about these things in the context of whatever we like doing day to day rather than getting uh you know overly caught up on the technology and the technical aspects which i think puts off uh some people that's

11:24Kanishka Narayan:that's a really key thing amazing and when you look at the tech landscape and you talked about the whole stuff you talk about llms do you think there's i guess what strengths do you think we have here in the uk so we don't have the you know the llms we have frontier labs but when you look at the talent, the research, the academic institutions, the entrepreneurial ecosystem we have there, where do you think the UK could really need when it comes to tech and AI? Yeah, I think all those fundamental inputs are fantastic. We have exceptional talent for out of the world's top 10 universities and an ecosystem that welcomes people and helps them build great outcomes.

11:59Compute, I think we're making huge progress on capital. I think we have a pretty good ecosystem and we're trying to do a lot more. All those inputs are good. But the big thing I'd say is we actually have strengths in the stack already. So if you look at chips, for example, not only do you have great strengths in photonics and compound semiconductors, but Arm, the world leader in chip design, is a British company started in a turkey barn in Cambridgeshire. We have a great set of inference chip companies that are building up as well. So as a chip layer, we're in a good place. In terms of models, I will point out that a lot of the Gemini team will be sitting in King's Cross as well.

12:30And then when you look outside language and other forms of intelligence, whether that's protein folding in science, or it's in what the team at Wave are doing or what Sathisia are doing, what 11 Labs are doing, what PolyAI are doing. We have a huge coverage in terms of really cool, great foundational models in some of these other contexts, as well as a great talent base here at Gemini and DeepMind as well. So the model context, we're in a good place. The fundamental game that I think Britain can really capture the big economic opportunity in is applied AI as well. And I've said before that I want Britain to be the world's adoption nation and right across the core areas of our focus.

13:07I think there's an opportunity for us to create tons of jobs in making sure that Britain is the place where there are lots of people in every part of the country working in manufacturing as a result of Britain being a robotics hub. There are great people in our services industry making sure that they're augmenting their work and their value where they are as well. So there's an opportunity to start.

13:26Kanishka Narayan:And what world do you think the tech ecosystem, the tech sector of all can play and drive the GDP growth to this country? Yeah, I think it is the main game in town when it comes to British GDP growth. And so it has everyone in the tech sector, and I come myself as part of it, both historically and now, we have a huge role to play at this. This is the primary driver of prosperity in this country. We've got to spread that prosperity. But the biggest thing to try and get to that prosperity in the first place is for all of us to think big, to talk positive and talk up this country and think about the huge price that's at stake.

14:00Kanishka Narayan:well I may as well thank you so much for joining me it's great to chat you've managed to rally the tech community behind you in a really short space of time and so yeah we're rooting for you and thank you very much super excited thanks Ed