What if the Prime Minister's AI Action Plan actually works?
Sir Keir Starmer announced yesterday that the UK government will be fully embracing the recommendations from a recent report on AI.
As an AI specialist for the last 15 years, I’m pretty upbeat about it – as are most people in the sector. It’s ambitious, long-term – and there are very specific objectives. If it goes well, this Action Plan will help keep the UK on the right side of the AI divide.
As ever, we need to see how these plans are actually implemented. But I can see one major risk: at the moment, this Action Plan is focused on what you could call the ‘supply side’. Making sure computing power is available; that data libraries are created and made accessible; that research centres are established. That’s all necessary – but is it sufficient? To truly become an AI-superpower, we also need to boost the ‘demand side’ too: that citizens themselves feel confident and capable to actually use AI in their personal and professional lives.
Will there be an ‘AI-divide’?
The promises of AI are enormous, and this Action Plan is right to emphasise them. Increased numbers of users will bring their own expertise to AI, driving increased time savings. Reports that AI assistants can reduce time taken for work by around 20% are fairly early-stage research, but in my experience, if an advanced expert begins to use an AI assistant as a ‘thought partner’, then this time saving could be closer to 80-90%.
Human expertise is crucial in realising the productivity benefit. Not only that, but more ‘consumers’ of AI will lead to further improvements.
But new technology is never picked up equally by everyone. Think about the advent of home internet and search engines around the millennium. There were those who knew how to ‘Ask Jeeves’ and those who didn’t. Those who did were able to access a whole new world of information; those who did not were left behind. Even today, disparities in internet access cause inequity of opportunities and earning.
The divide between those who take advantage of new technology and those who do not happens all the time. But with a technology as powerful and transformative as AI, the consequences of such a divide could be huge. We must not simply replace the old ‘digital divide’ with a newer, even more dramatic, ‘AI divide.’
The Action Plan is full of laudable proposals about ensuring public services and research institutes can benefit from AI. But there is less about ensuring that people – ordinary citizens, business employees, staff – actually pick it up and benefit from it.
There are AI Makers and AI Takers – we need to master both arenas. We must become Makers who build, and continually improve AI models, start companies, and develop new applications.
But we must also drive adoption so organisations become effective AI Takers (leveraging AI as a service), and citizens become confident users, who begin to automatically reach for AI tools as a thought partner (what I call a ‘walking stick for the mind’. My colleague Richard Townsend calls it the ‘motorbike of the mind’. Perhaps he’s even more ambitious than I am).
We already know there is a lot of fear and nervousness about AI taking people’s jobs or running rogue. (Some of it driven by inaccurate scare stories and Hollywood movies, in which AI is always the dangerous bad guy).
There is still too much pessimism and fear about AI displacing jobs. Think about self-checkouts. They put a lot of cashiers out of their positions, but they employed a lot of the engineers who designed that new automated system. AI will bring about the same. Careers in data, AI, and machine learning will continue to boom, with demand for skills only rising as an explosion in AI users demand that the product gets better.
The role of leadership in AI
That divide is not just a global or national one, it exists within businesses too. Your AI Makers, the people building and providing models, are not the only ones who will require new skills. Users of AI across your business, finding efficiencies and improving their impact, need the right skills and knowledge to effectively and safely adopt AI.
That is because, in my experience, AI adoption works from the top down. Leaders must be invested, so their people can follow confidently, with a shared vision and the appropriate guardrails. If I had five minutes with our prime minister, I might ask him about his personal journey with embracing AI. As a leader, his buy-in plays a part in setting the direction on a national scale, just as your CEO’s buy-in affects the direction of your business. If your boss is scared of AI, you’re unlikely to use it, right? If they champion it, understand it, and set out clear directives for its valuable and secure usage, you’ll be a lot more keen.
If there is one thing that every CEO thinking about AI adoption should bear in mind, it’s this: make a clear case, and a clear mission, about why and how the company is going to use AI. Make sure people don’t approach it with fear or apprehension – but rather excitement and confidence.
Make no mistake, the changes coming will be immense.
Twenty-five years ago, did we expect self-driving cars to come along as fast as they have? For that matter, what kind of phone were you using? Even in the 15 years I’ve been working in AI the improvements in generative AI especially have been stunning. How many of you had even heard of ChatGPT just two years ago?
Once we get the product better (which we will) and the experts confident in using it, the sky is the limit. That’s everything from tackling global challenges down to very small life hacks designed around one person. If UN bigwigs use AI as a true thought partner, novel ways to tackle war and climate change will emerge. But it might also be a head of marketing at an SME freeing up 2 days a week by using generative AI to co-author reports. And everything in-between.
To get there, the demand side must keep up with the supply. That will take a mindset shift, and buy-in from bosses. More widely, I imagine one day soon we will require some kind of AI literacy for all – in schools, apprenticeships and beyond – so that the benefits of this remarkable technology are shared as far and wide as possible.
Check out our learning solutions in AI & Machine Learning to start training your organisation and yourself for the AI revolution
