How AI could change the course of UK productivity: Read our article to find out more

The UK's productivity challenge

Productivity is now well behind where pre-2008 trends should have taken the UK by 2024, with current levels currently 24% lower than they would have been if pre-crisis trends had continued. The gap has far-reaching implications for the economy as a whole and individual businesses alike. While factors such as Brexit and the COVID-19 pandemic have unarguably contributed, the underlying productivity problem is an issue that we cannot assign to any single global event. But, unlike Brexit or the impacts of a pandemic, we can all do something about it.

The AI revolution as a possible solution

AI could be the key to unlocking significant productivity gains and improving UK competitiveness – but to do that we need to enable great innovation at pace if we’re to stay ahead of the game. We have known for some years that AI could add up to £232 billion to the UK economy by 2030, equivalent to a 10.3% increase in GDP. This impact is expected to come from productivity gains and consumer-driven demand for AI-enabled products. Generative AI alone could potentially double the UK's long-term growth rate from 1.6% to 3% annually, potentially adding more than £700 billion to annual GDP by 2038.

Industry leaders Workday claim that using AI could save big businesses in the UK close to 8 billion employee hours per year, with individual business leaders and their employees saving up to a total of almost 2,000 hours annually if AI were used to its full potential. Financially, AI could unlock £119 billion worth of productive work each year across UK large enterprises. For a single large business with over 10,000 employees, this could equate to an average of £110 million annually. A recent Goldman Sachs report points to robust evidence for productivity gains of around 23%, while anecdotally some companies report a figure close to 30%.

These numbers might seem like hyperbolae, but the reason the calculations stack up is because AI can transform productivity in such a wide range of ways. While the industrial revolution almost solely impacted labourers, the AI revolution can and will impact jobs of every type, in every sector. It can automate mundane and repetitive tasks, freeing employees to focus on more meaningful and high-value work. AI-powered tools can convert complex datasets into intuitive, actionable insights, enabling faster and more informed decision-making. AI can lead to better quality and more personalised products, all while extending human capabilities through a combination of automation and human-AI collaboration.

Therein lies the challenge

The folks who used to sell copies of the Socialist Worker on my university campus wanted more students to believe that capitalism was about pursuing wealth and growth for the sake of, well, wealth and growth. That’s probably true for some people, but it’s not why we care about productivity and growth, it’s probably not why you care, and it’s definitely not why governments have to care about it. Governments are, however it might sometimes feel, all about the people – whether it’s as a point of moral principle or for the sake of their re-election. Your government knows that the state of the country’s economy and your day-to-day life are inextricably connected. Growth leads to higher output, which drives demand for employees, which drives employment and wage growth, which improves your standard of living and your purchasing power.

For any Government, it’s also a balancing act: growth can also drive inflation if demand outpaces supply, and that can pull the cost of living in the wrong direction. Get the balance right and you reduce absolute poverty, improve lives and stabilise society – as well as generating tax revenue that can be reinvested in services that benefit everyone in both the short and long term. Things like education, health and infrastructure.

In this context, AI could be a pretty neat fit: well-placed to grow productivity so you can keep pace with demand and dodge the inflation risk. But while the potential of AI to improve productivity is hard to overstate, that means there are inevitable challenges and growing pains ahead.

AI is going to generate new opportunities but it will also affect the job market in ways that will negatively impact people you know and care about. Only a few weeks ago one of our team learned that a relative in his early 50s was being made redundant in an industry where the increasing use of AI tools is already making significant inroads into the need to employ professionals in creative roles. This is the reality of where we’re heading.

Resistance won’t put the genie back in the bottle, and neither should it. The problem lies in how we get to an AI-enabled future state to unlock the best possible benefits with the least pain for the generation caught in the middle (which is most of the workforce): anyone who is neither senior enough to be an irreplaceable, c-suite decision-maker, nor early enough on their career paths to leap at the chance to retrain.

A way forward

First, we have to work out how to release all that AI potential, rather than wait and see how everyone else does it and risk falling behind. And that will require vision, concerted effort, strategic investment, and a focus on responsible AI deployment across all sectors of the economy.

Accenture estimate that up to 62% of the UK workforce may need reskilling to take advantage of AI-driven changes, while less than half of employers have increased training in generative AI. Businesses will need to increase their investment in AI technologies and focus on scaled deployment to realise the full benefits. There's also a pressing need to build trust in AI technologies and ensure their responsible deployment.

If Engine were playing policy wonk today – and that is honestly one of our favourite pastimes – we would argue there are five things on the big AI readiness ‘to do’ list, if we want to give the UK it’s lost ‘edge’:

• Strategic investment: The government - and businesses - should prioritise investment in AI technologies and skills development.

The previous government announced a £1.1 billion skills package targeted at future tech including AI, with another £80 million for AI research across regional hubs, and a doubling of funding for the Turing Institute. More recently the 2024 budget included an £800 million investment plan to ‘leverage’ (classic AI speak) technology. But announcing spending is (a) a lot easier than landing the money where the action is and (b) meaningless if no one has decided quite what to do with it. Also (c) governments have a worrying tendency to ‘re-announce’ existing money, which means the ‘new’ investment isn’t really coming at all.

We also need to ask ourselves if the scale of investment is fit-for-purpose: the government’s £1.1billion is less than half of the sum that Microsoft have announced as their own spending on UK AI over the next three years – mainly to expand their maxed-out datacentre infrastructure. And for most small and non-profit organisations, the idea of ‘new funding’ is illusory: your skills and tech buying current budget needs to be reprioritised, fast.

• Regulatory framework: The UK must have a world-leading regulatory framework that encourages leading edge innovation while ensuring responsible AI use.

The UK has a post-Brexit opportunity to remain apart from the highly risk-averse approach the EU are rapidly adopting, which means it could become a world-class centre for leading edge AI development – safe from the US government tendency to peek at your data at will, and usefully outside the controls imposed by the EU.

Instead of EU angst and rest-of-the-world data privacy risks, the UK has already established principles for the design, development and use of AI, a hub for innovation and advice, and a function designed to bring coherence to any regulation. The only worry is: that was all done by the last government. It will be hard to resist the temptation, as a new government, to go further once everyone starts over-thinking things. It’s hard to hold your nerve and not develop policy when developing policies is the day job.

• Cross-sector collaboration: We need to foster collaboration between academia, industry, and government to drive AI innovation and adoption.

There’s a partnership fund, the new DRCF Digital Hub is cross-regulatory… but government is pretty entrenched in siloes and that makes it harder for business to work across different frameworks too.

The impact of academia is an easier win: one of the consequences of amazing generative AI is that evidence and data-driven analysis is everywhere. It’s why, instead of writing a blog full of half-considered AI musings, I can use Perplexity and find facts and information to prioritise, shape and evidence a five-step plan. And we’re seeing this everywhere: AI brings knowledge as a free gift with every purchase.

My advice is that this one will be on us, not on government, unless you’ve missed the point and think we have time to perfect the art of civil service reform first. Government is huge and they have to have an organising structure: it’s our job to disrupt our own feed. Talk to other businesses and colleagues in other sectors and think about whether what they’ve done can work for you. Government should focus on where innovation is most needed to improve outcomes and least likely to happen organically because investment is scarce. I’d start with letting everyone do what North Yorkshire has done for their social workers.

• Skills development: Everyone needs to implement comprehensive programs to upskill and reskill the workforce, preparing them for an AI-driven economy.

The civil service are being offered nine new training courses, we’re told. Then there’s the hilariously tiny upskilling fund for SMEs announced by the previous government. Some of these small-scale training courses will make a difference, but two more things need to happen to develop skills: CPD was never just about going on a course.

First, there needs to be serious adjustment to the education system to prepare children and young people for a different world of work. The supply of vocational, professional and academic skills needs to match the change in employers’ demands. That’s not about sending a civil servant on an AI ethics course – which can be achieved cheaply using a few shared slides imho. This is about computer science not being a core subject. It’s about how many graduates of which flavour are sent out in the world by our colleges and universities. And, just like we’ve had to with calculators, mobile phones and other previous tech innovations, it’s about the teaching workforce embracing the use of AI both inside and outside the classroom as it becomes a more and more intrinsic part of students’ everyday lives.

Second, for all employers, it’s about what you give your team to empower them to learn and grow their own productivity – and if you’re one of the ones that switched off ChatGPT rather than tell people how to change the privacy controls, know that I’m looking at you right now. It’s like hiding the balance bike in the shed until everyone has learned how to ride like a pro at the velodrome – but then not teaching them to do that either. Show people what is possible, be a role model, create safe spaces to Try Things. Have a policy, put some guardrails in place, allow change to be driven by what makes your people happier and more productive. Let the flowers bloom.

• Responsible AI Deployment: Yes, we have to ensure AI is deployed in a way that augments human capabilities rather than simply replacing jobs, and that it does no harm.

Regular readers of the Engine blogs will know this is one of our favourite themes: you need to help your teams adapt, be mindful of their inevitable anxieties, and watch for bias. Nato’s AI strategy is a stunning example of how achievable this is at scale, and for our own solutions it’s why we deliberately build our tools to keep the human in the loop, keep data secure, make facts verifiable and watch for environmental impacts.

But we know that most of you haven’t adopted an AI strategy or even a generative AI use policy. We know this because lots of you filled in our survey: only one in four told us they had a plan, and only half have a usage policy. So we have to admit there’s more we can do. If we don’t, we can’t complain when the government despairs, and then regulates us into submission.

What happens when we lose the wrong people from the workforce

The UK also has to face up to a particular problem that compounds the productivity crisis: the increase in economic inactivity across people in their 50s, especially women. Many of those leaving work before retirement are citing a depressing mix of health issues, the menopause specifically, caring responsibilities and – for some key professions – the impact of public sector pension caps on their earnings. And worrying about the potential impact of new technologies AI on your job isn’t going to help. It might make more people think they need to jump before they’re pushed, taking all their experience and knowledge with them.

The previous government looked for answers in the tax system, but the reality is that it isn’t just about the money, this being the last generation that could generally afford to buy family homes, pay off their mortgage and live off their pensions. They are literally sick and tired. And it is just us: inactivity for this group has remained stable in Germany and fallen in France, even though they had the same pandemic and slower vaccine roll-out, so this isn’t just about long Covid, and it isn’t a global trend.

The previous government belatedly adjusted the pension cap, but nothing else about this issue has changed since the problem materialised in the pandemic, and we’re losing experienced teachers, doctors and a host of other professionals as a result – as well as driving down diversity, which damages business performance.

Yes, some of this is about the current state of the NHS, how that impacts our health, and the cost of alternative options for care (and having to step in to provide affordable grandchild care), but it’s also about how employers adapt to retain staff who they have invested in developing for decades. Somewhat obviously, here’s what doesn’t help: inflexible working patterns including excessive (and poorly communicated) restrictions on working from home, and everyone having to spend time on the parts of the job that are not what motivate you to get up each day. Therefore, inevitably, what can help is more diverse leadership teams that actively improve inclusion, encouraging flexible working patterns and job shares, and giving people what they need to perform at their best. And that’s definitely where AI can help.

Tech for all

This isn’t a marginal issue: 3.5 million people aged 50-64 are now classified as economically inactive. The economic inactivity rate for women aged 50 to 64 is significantly higher than for men of the same age at 31% versus 23%. If you’re not convinced about the scale of this issue, read this: Economic labour market status of individuals aged 50 and over, trends over time: September 2024 - GOV.UK.

This might not be the obvious target group for retraining, but technology – including the integration of AI tools like simple, secure, GPTs - can help address some of these issues. It’s unarguably more expensive to keep increasing our investment in shiny new teachers just to keep pace with turnover, for example, so isn’t it a bit weird, really, that we let great teachers leave our schools rather than helping them find a job share partner and/or giving them access to some basic help that alleviates their admin load?

It's hard to imagine we can catch up and overtake our European neighbours on productivity in general while we’re letting the economic inactivity problem fester – even if the EU has locked down some AI tools while the UK hasn’t. I hope it’s not because I said the word menopause – is it? Finding the topic a bit awkward would be a terrible excuse for inaction. The Boston Consulting Group concluded that companies with a diverse age demographic can be as much as 25% more productive. Deloitte reports also pointed to improvements in engagement, retention and lower recruitment costs. It makes your workforce more resilient. That’s right: having older staff makes you better at change. They like tech too – they just don’t like bad tech, usually because they’ve wasted time on it before.

AI tools like private GPTs – our particular area of interest at Engine – aren’t the only way to keep experienced staff engaged and productive using AI. AI-powered chatbots can initiate check-ins with employees, offering a platform for them to express concerns or provide feedback. AI can facilitate personalised learning platforms, enabling workers to acquire new skills at their own pace. Implementing wellness devices and health monitoring apps can help employees manage chronic conditions effectively. But AI can certainly automate routine tasks, allowing your most experienced workers to focus on higher-value work that makes best use of their acquired experience and knowledge.

What's stopping us?

Despite the obvious benefits, many organisations remain hesitant to adopt AI for reasons that are totally relatable. This isn't the industrial revolution, so the issue isn't as simple as a few luddites chucking their sabots* into the machinery. Unlike the Spinning Jenny or the first steam locomotives, you can't ‘see’ how most new AI tools work and for many that's unnerving. But by not providing AI tools to make jobs easier and the tech more familiar – including, or maybe especially, for older workers – companies risk losing valuable human resource and continuing to preside over a productivity decline.

Here are some Engine New Years Resolutions to kick us off:

1. Advocate for investment: When the government inevitably organises roundtables and consultations, or when your company looks at the budget, actively make the positive case to invest in the right tools for transformation and only adopt controls where you need to. Respond to ‘no’ with ‘why not?’ Hold the door open.

2. Try new things, and do it out in the open: Use a GPT to create an email to your staff, show them how you did it, encourage them to use AI tools too and be honest about how and what they did. This can help bridge the gap, especially for anyone hesitant about AI adoption.

3. Start today: Don't worry that you're behind or that it's not for you: get started. Ask your teams where the pain points are and we'll help you work out how to alleviate them. Try New Things. Read about it, learn about it. Only 15% of workers aged 45+ are currently using AI tools at work - and only around 6% of companies have started to use AI as standard. If you’re reading this, you’re probably ahead of the curve.

"You can't be that kid standing at the top of the waterslide, overthinking it. You have to go down the chute." - Tina Fey

‍

* Sabot: a kind of clog. Sabot + machine x chucking = sabotage.

‍