For most of the modern political era, economic debate has revolved around one central question: how do we create more jobs?
But what happens when technology begins reducing the need for human labour just as our population is ageing and demand for care, health and support is rising sharply?
Artificial intelligence is already reshaping parts of the economy at extraordinary speed. Entry-level legal work, coding, administration, customer service and research are all changing before our eyes. At the same time, Britain is growing older. More people are living longer, often with complex health or care needs, while birth rates fall and traditional career structures become less stable.
Some see this future and respond with fear. Others retreat into nostalgia, promising a return to a world that no longer exists – if indeed it ever did.
Liberals should do neither.
This moment demands something far more ambitious: a redesign of the economy around human flourishing. Because the real question is not simply how many jobs exist. It is whether people are able to develop their potential, contribute meaningfully, live with dignity and freedom, and participate fully in society throughout their lives. That, surely, is what Liberalism has always been about.
Liberalism at its best is not an ideology of atomised individuals competing endlessly in a market. It is a philosophy of human liberation. It asks how we remove barriers that prevent people from becoming who they are capable of becoming.
That means equality of opportunity. It means lifelong learning. It means decentralisation of power. It means freedom from poverty, insecurity and ill-health. And it means recognising that worthwhile work matters not only because it pays the bills, but because contribution, purpose and dignity are fundamental human needs.
For too long, our economic system has treated people primarily as units of production and consumption. GDP rises when forests are destroyed, when prisons are built, or when pollution creates costly clean-up operations. Yet none of these things improve human wellbeing.
If AI dramatically increases productivity, the central political challenge of the coming decades will be this: who benefits? If the gains flow only to a small group of technology companies and asset owners, we will see rising inequality, social fragmentation and political extremism. But if those gains are shared, AI could help create a society with more freedom, more creativity, more care and more time for the things that make life meaningful.
That requires active choices.
First, we need to treat care as economic infrastructure. An ageing society will require many more people working in health, care, rehabilitation, mental health support and community resilience. These roles should be among the most respected in society, not among the most undervalued. AI should help remove bureaucracy and drudgery from these professions, freeing people to focus on the deeply human aspects of care that no machine can replace.
Second, we need a genuine lifelong learning revolution. In a rapidly changing economy, education cannot end at 16, 18 or 21. Every citizen should have access to lifelong learning and retraining throughout their lives. Liberalism has always believed in the development of human potential. In the AI era, that principle becomes even more important.
Third, we should guarantee young people a pathway into adulthood. One of the gravest dangers posed by AI is the hollowing out of entry-level work that traditionally allowed young people to gain experience and confidence. We should create a Youth Guarantee offering every young person access to paid apprenticeships, technical training, civic service, green jobs or care placements. No young person should be left feeling economically unnecessary.
Fourth, we should ensure that the gains from AI are widely shared. If AI systems are built on humanity’s collective knowledge, culture and data, then the benefits should not accrue exclusively to a handful of corporations.
This isn’t pie in the digital sky. The UK’s existing Digital Services Tax already raises around £800 million a year despite applying only a modest levy to selected digital revenues. An expanded AI Dividend Fund could build on that model through a mix of taxes on exceptional AI-driven profits, stronger action against offshore profit-shifting, and fair taxation of capital gains arising from AI-enabled wealth creation.
The point would not be to punish innovation – quite the opposite. Liberal societies thrive on innovation. But when technology transforms the economy on this scale, the public should share in the upside too. That money could support lifelong learning accounts, youth opportunity programmes, social care, community resilience and universal basic services that expand real freedom and security for everyone.
And finally, Liberals should lead the conversation about what the economy is actually for.
The purpose of the economy is not simply to maximise output. It is to help human beings live good lives. A society in which people are permanently exhausted, isolated, insecure and unable to fulfil their potential is not truly prosperous, no matter how high the GDP figures climb.
We should aim higher than that.
The AI revolution is coming whether we are ready or not. The question is whether we allow it to deepen inequality and insecurity, or whether we shape it in service of human freedom and wellbeing. Liberals should have the confidence to answer that question boldly. Not with fear of the future, but with a vision of a society in which technology helps liberate people from drudgery, expands opportunity throughout life, strengthens communities, and gives every person the chance to flourish.
That is not only an economic vision. It is a Liberal one.
* Dr Roz Savage is the Liberal Democrat MP for the South Cotswolds.
Follow @libdemvoice.org on Bluesky
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
Martin Wainwright's Guardian obituary of Michael Meadowcroft
Evaluating different electoral systems for the House of Commons
Brexit and summer holidays
The shame of Liverpool Labour councillors taking their allowances but not turning up to do the work
Peter Martin
Peter Davies
David Wright
11 Comments
Why? Why is it inevitable that we let the planet-destroying brain-rotting plagiarism machine take over entry level jobs? Where will the people filling senior jobs come from if there’s nobody starting at entry level?
I know I’m repeating myself here, but enforce copyright law on the stolen training data these “AI”s run on, and fine them for the massive damage they are doing to the environment to fuel their glorified autocorrect.
There is a case for the mass scale gen-AI LLMs being pushed by the industry at the moment, but it is not a Liberal one, because it necessitates destroying the environment, intellectual property rights, and individuals’ ability to think and act for themselves while transferring both wealth and the ability to earn it over to a small number of foreign interests.
As Jennie says, any credible Liberal response would involve muscular enforcement of environmental, data protection, and copyright law, along with market controls to enable users to decide how much AI they want and need in the patforms, services and products they are using. That is how we ensure it doesn’t deepen security and inequality.
“The AI revolution is coming whether we are ready or not.”
A bunch of really expensive chatbots that give vast quantities of mediocre output is hardly a “revolution”, any more than virtual reality or blockchain or indeed the previous “AI” hype cycles were. Lots of talk about what “AI” could do to distract from what it does do being both highly unethical and not especially useful.
I am far more worried about the economic shockwaves from the impending collapse of the current AI bubble than “what if they make a slightly better chatbot”.
(The general recommendations of the article around who should be benefiting from the economy, taxing tech multinationals heavily, etc. – sure, all good. But that’d be good any time)
It is not AI we need in the care sector, it is better pay. At least parity with health workers and deliverers of Domino’s pizzas.
The phrase “AI is going to reshape our economy” should be understood as a THREAT.
The real question is what we should be doing to address and prevent the ongoing parade of scams that exploit negative externalities to line already rich tax-evading fascists pockets with yet more money, while leaving government (& therefore ordinary people who DO pay their taxes) to clean up the mess they leave behind.
Cryptocurrency, NFTs, Blockchain, AI – all these are just extra faces on the same scam, funnelling money to assholes like Thiel and Musk while poisoning and impoverishing the world.
The best thing anyone can do to an AI datacentre is to burn it down. It’s even less emissions than letting it run!
One benefit of my professional membership is being able to attend meetings on AI including last week’s at the Institute of Physics in London where two of the speakers were from NPL. A comment from one of the NPL speakers stuck in my memory, that developers are starting to look now at Small Language Models rather than Large Language Models. The difference is that if you want to make an AI to perform, say, entry-level legal work, it makes no sense to scrape the whole of the internet for training data. A smaller model that can be trained on legal documents alone turns out to produce better results with fewer hallucinations. Let the American tech bros have their LLMs; we do not have to follow them or build the gas-guzzling giant data centres that the LLMs need.
Sorry but this article seems heavy on wishful thinking and light on how or why. So far, the evidence that generative AI actually delivers significant productivity improvements is pretty weak. What we have is lots of hype from the loss-making tech companies, the lobbyists on their payroll, company bosses looking for an excuse for lay-offs, and Governments desperate for some magic sauce to make their future budgets add up.
Any liberal who believes in this technology needs to answer the following questions:
1) How will we satisfy AI data centre’s appetite for energy without losing the battle against climate change?
2) Ditto the demand for scarce fresh water?
3) How do we protect the livelihoods of artists, writers and musicians, and return to them what has already been stolen?
4) How do we prevent AI from accelerating digital exclusion?
5) How do we avoid further loss of digital sovereignty?
It’s not “responding with fear” or “retreating into nostalgia” to ask those questions.
Labour has already shown an alarming willingness to throw the UK’s successful and long-established creative sector under the bus in the hope of currying favour with tech companies. We shouldn’t follow them.
In fact, this article works better if you just remove all the references to AI. For example – “An expanded Dividend Fund could build on that model through a mix of taxes on exceptional profits, stronger action against offshore profit-shifting, and fair taxation of capital gains arising from wealth creation” sounds good to me….
A very good crop of comments today (apart from burning down DCs, why not lawfare them into penury, which is legal?). Why do we not see these arguments in the mainstream media?
Many of the comments are the very objections I had to Labour’s Industrial Strategy (“let’s make the UK even more friendly to US private equity and Big Tech and increase dividend payments to offshore – also destroy 8 sectors in the economy while we are at it”). They handily wrote it all down. Later it emerges that someone with a VC Silicon Valley background wrote it for them. Also they adopted the Tory “freeport” idea and re-branded it AI zones, essentially democracy free zones where pirate can operate. Only we pay for the water and electric that these places need.
@Ruth: there is a vast truth behind what you say. Like the perennial arguments about skills shortages. With liberal supply-and-demand it reveals a wage shortage. Raise wages and skills suddenly appear. They did in the industrial revolution. And in my family history. Lower wages and skills drift off. Civil servants write lots of reports about that.
Thank you to everyone who has engaged with this article – including those who disagree with parts of it.
I absolutely agree with those who say meaningful work matters deeply. Work is about far more than income: it gives people dignity, purpose, connection and a chance to develop their potential. Any liberal response to AI has to keep that firmly in mind.
At the same time, AI may disrupt the economy faster and more profoundly than previous technological shifts, particularly alongside ageing populations and growing pressures on public finances. The question is not whether this change is coming, because it already is. The question is whether society shapes it wisely and shares the benefits fairly.
In an ideal world, governments would already be getting out in front of this with long-term thinking and serious public debate. But I am simply not seeing that happen at anything like the pace required.
For me, this comes back to a very liberal question: what is the economy for? If technology can reduce drudgery and expand human freedom, that could be a wonderful thing – but only if we build an economy that helps people feel more empowered, secure and able to thrive, rather than more disposable.
I certainly do not claim to have all the answers. But I do think liberals should be engaging with these questions now, not later. And I thank you for doing so.
I’m a software developer by trade and I’ve been told my job will be replaced by computers “in a few years” since I graduated in the mid-90s. Still waiting. We’ve been here before with technology many times – the programmable loom will kill off textile jobs, robots will kill off car factory jobs, and yet there are probably more people employed by those industries than ever before.
LLMs are not the future of artificial intelligence, and the tech bros pushing them (who were previously pushing fads like blockchain, NFTs, and “the metaverse”) are running out of money fast. Companies are realising that the tech isn’t providing significant improvements – indeed, some tech companies are suffering increased outages and revenue loss because of “AI” generated software in production. Some have had to re-employ workers to check the work done by their “AI” replacements.
This is one of many “AI” bubbles over the years and is going to burst, hopefully soon because the “AI” companies are stripping the world of hardware for datacentres that haven’t been built yet, pushing up prices and trashing supply chains. Plus, as mentioned by cim, the economic effects of that collapse is quite likely to be of significant impact.
We need to learn to view paid work differently. It is not a right but a gift by society to earn income and other things. We will need to learn to live in a world where work is intermittent, valued and earned. For those less fortunate we need to create a society that can flourish without the necessity of paid work.