AI is happening. We are not going to stop it, and we shouldn't pretend we can.

But we can manage the economy that AI will reshape. And we must.

In this video, I explain what it means to manage the AI economy through regulation that makes AI pay its full resource costs and through an investment-led programme to cut inflation structurally while creating the jobs AI cannot replace, most especially in energy, housing, skills, transport, and care.

This is not about surrendering to market forces. It is all about the government staying in charge.

This is the audio version:

This is the transcript:

I've done a number of videos now talking about the problems that AI could create in this world, and in particular in our economy.   I have argued that AI can, and probably will, cause inflation, and it can, and probably will, cause unemployment, and these two factors, in combination, could result in the imposition of higher interest rates from our central banks, or the imposition of austerity by our governments.

None of these are very attractive options, let's be honest. AI does not look like a panacea for well-being as a consequence. So what can we do? How can we manage the AI economy in that case, given that we know AI is going to happen? That's the question I want to look at in this video.

AI is happening. We're not going to stop it. Let's not pretend we can. Let's not pretend we even necessarily want to. There are some benefits in this technology; let's not pretend otherwise. But we do know it can cause real problems. We know that at present, very few companies can see a positive use for AI, in the sense that they cannot see new products and services that they will generate as a result. Instead,  they see a cost saving for them by making people unemployed. The risk is to real people as a consequence.

We also know that AI is an inflation risk. Why? Firstly, because it demands more electricity, and that will push our prices. Secondly, because it demands more water, and that could push up prices. And thirdly, because it could create shortages of IT equipment, which in the  short term might have the biggest price implications of all. It's threatened that IT price increases in 2026 might be as high as 20%. In   that case, there's no good news for inflation in this.

That combination of high inflation and increasing unemployment is, however, something that is said to be impossible in orthodox economic models. The so-called Phillips Curve rules it out as a possibility. But what is clear is that it is now entirely plausible, and the reaction could be higher interest rates from central banks as a result, because that's the only reaction they ever have when they see inflation rising, and we could see austerity from governments because they will say the inflation that is being caused, and the higher interest rates they're suffering, must require a cutting government spending.

Frankly, this looks like a completely ridiculous response, but it's one that is entirely predictable. The outcome would be perverse because it would intensify recessionary pressure. It would worsen household incomes at a time when  people will be losing their jobs. It will damage businesses   outside the AI sector, and that can be of no benefit to anyone, and it will discourage long-term investment in anything but AI, and again, there's no chance that this will improve well-being as a result.

In fact, what we'll get is something like a disastrous economy with high interest rates, low government spending, high unemployment and high inflation, which is a nightmare scenario unless we think ourselves out of the false choices we are being presented with.

The false choices are very clear.

We could be told that we have to have inflation control, or full employment, and   central banks will always go for inflation control.

And  we'll be told we have to have austerity, or we have to increase deficits, and as deficits aren't   allowed according to neoliberal thinking, austerity must come our way.

Of course, all of this is false framing. It is an old economics for an era long gone, facing the challenge of a new technology in the form of AI that changes all the economic relationships that we know about, and which therefore requires us to reframe the whole of our thinking about the economy, in particular rejecting the household analogy that has been used to underpin the false agendas that have driven austerity for the last 15 years.

A wise government would see that.

A wise government would reject the choices that orthodox economics will present to it if AI moves in the direction I predict.

But we aren't sure as yet that we have wise governments.

So what we need to do is think about what should take place in a world where we know we face risks from AI.

What should we do to manage the economy that we are going to have, in other words? That's what I'm really interested in. And the first thing I'd say is that we should be challenging the threat at source.

The threat to our economy is going to come from AI just as much as it creates opportunities for some.  That threat from AI has to be dealt with by regulation. It can't   be dealt with by the standard techniques of inflation management that we've got now, because they won't work because there are no relationships left which they are meant to manage.

Instead, we have to challenge AI directly by charging it for the whole cost of its electricity. In other words, if AI is going to push up the  price of electricity, the only person who should suffer the price increase is the AI industry.

And if the entire cost  of water is going to go up because of the cost of AI, then the entire increase should be borne by the AI industry.   And if the stress that this is going to create is going to reduce demand for AI data centres, we shouldn't be permitting so many. By implication, we will have to slow down the expansion of the AI industry to control the inflation consequences that it might give rise to.

In other words, we should not, as our government is doing, and many other governments around the world are doing, are saying, "  Please come and do AI at any cost." We should be saying, what is the cost?   And let's tailor a transition which manages those costs within sustainable resources. That's the key point now.  AI cannot run the economy into the ground; it should instead be managed. The government needs to stay in charge.

But that, maybe, is not enough. Those measures cannot, by themselves, probably sustain an economy if we are really going to face the AI challenge. What we must do is actually look at how are we going to adapt to this new world where there will be fewer of the conventional jobs that we've had?

Now, let's be honest, we've been through this situation before. We've seen  half a million miners lose their jobs in the UK at one time.

We've seen half a million people who worked on the railways at one time lose their jobs.

And of course, the entire typing pool that once provided masses of female employment in this country has simply disappeared.

So let's not pretend we can't manage changes of employment practice in the UK: we can. But we can only do so if we expand alternative capacity and training and create the resilience within our economy which manages this process of change to create jobs and reduce inflation pressure over time. That's critical, and that's what we need to do.

Let's look at how we can do that. Firstly, we need a range of investments that cut inflation, and I do mean a range. There is no one solution here. So, apart from the fact that we need to regulate the amount of power that AI can use, we also need to look at how else we can cut the costs of energy, because we can.

We can reduce consumption, for example.  We could do home insulation and retrofit of houses. Not only is this immensely valuable with regard to job creation,   and we do need jobs, it also cuts consumption, and that is key to managing the cost of energy demand. Therefore, we could reduce energy demand, we could reduce the exposure to imported energy shocks, and  we could lower household costs, and we could therefore lower   inflation pressure simply by doing this one task, which I've been promoting for 20 years now. Literally create what I've always called a 'Carbon Army' of people to go around the country and do home insulation and retrofit, and we will produce resilience within our economy and lower inflation pressure.

We could also change our energy supply because we need to do that as well.  Renewable investment is critical at this moment. That's not just reducing demand. It's actually changing the way we generate.   And we need to change the grid to match, of course. We all know that the National Grid is unfit for purpose. There is a need for a massive investment in the National Grid. Not one that is driven behind the market, but one which anticipates the market, and the new market is for renewables; therefore, the grid must change to suit, and the government must be providing the funding necessary to deliver this.

I have explained how the government could do that in other videos. It could change our savings systems via ISAs and our savings systems via pensions to require that both are invested in infrastructure in the UK.  The money to fund these investments could be provided if only the government changes the tax incentives on savings in this country.

I believe £100 billion a year could be released in this way to literally invest in the programmes I'm talking about here.

But the consequence of investing in the renewable energy industry that I'm now talking about is that we would reduce our vulnerability to  fossil fuel price spikes, those things that have caused so many problems for us in recent years.   And we would create a more stable cost base across the economy with AI paying full whack for its costs within that structure to make sure that we are not subsidising it.

What's the third form of investment we should make? Well, we should be investing in public infrastructure and supply.  We need to rebuild our public transport because it is too weak. We know that that's necessary.   The government is talking about spending £45 billion on Northern Rail at present, and it has got so much it needs to do in the South of England as well, plus Wales and Scotland. All parts of the UK need investment, and again, I have found the money to do that. It's available if we want to, so long as we can train appropriately qualified people to do the job. But this is critical now.

It's also vital that we look at how we create supply chain resilience. There   are key products and services that we buy from overseas, which are not resilient. In particular, foodstuffs. It's ridiculous that we are now in the situation where we grow so little of our own food supply. Let's just imagine for a moment that we are going to have those AI data centres producing vast quantities of heat. Why aren't we going to put next to them the  greenhouses that could eliminate our import of, for   example, many short-term foodstuffs like salad crops? Why aren't we putting next to them the greenhouses where we could grow tomatoes all year round and so much more?

All of this could be done. We could improve our resilience. We could do this in other sectors of the economy as well. The point is, do this structurally give grants to help it happen. That's what a proactive government should now do. There is no way on earth that we should now be allowing an AI data centre to exist without it adding to some other form of product supply within our national economy.

And we need to also invest in housing because that will also cut inflationary pressure. Housing repairs are necessary. They reduce waste. They reduce ill health. They increase productivity. They reduce the pressure on pricing as a consequence. And new construction expands real capacity within our country, which is necessary because our population is still rising, and housing costs are a core inflation driver. This is why delivering this programme within our social housing system is a priority because people need stable housing in the long term at an affordable price that they know will be guaranteed for a lifetime, because that's what people need to live in community.

Finally, and this is the last of these investments, to ensure that we can control inflation,  we need to invest in skills and care. We need skills investment,   and in fact, that is something that the IMF is literally saying in the last week. They're saying it is key to the economic transition to AI, and that is true in this country, but the government has to deliver this because the investment will be in jobs for which there is limited demand at present but for which there will be much demand in the future if only we can get ahead of the game.

So the time has come for the government to take the initiative here, and not markets, and we have to do the same with regard to the care economy. The care economy is fundamental to our future well-being because it is going to be about care when it comes to employment, because AI can take away the drudge jobs, but people will want to work, and the best chance for people to work is to supply them with jobs that AI cannot readily replace, and  care is right up on the spectrum of jobs that can't be replaced by AI alongside, in my opinion, education, because like it or not, people need people to teach them what to do.   All of this will build resilience and social well-being into the real economy and provide the jobs that people want, which will be rewarding, which actually would be a positive development from AI.

All of this will require a greater depth of economic understanding by a government; let's be clear about that. I'm afraid to say that very often our governments have worked on very simple metrics. For example, they talk about unemployment in the UK as if it is a single number, which is complete nonsense. Unemployment varies by region, by sector, by age, and also by disability and neurodivergence type as well, and this last point is particularly key when it comes to young people because we know they're suffering high rates of unemployment because of these issues.

Blunt policies are not now fit for purpose.

Poor data can't now let us manage the risks we face.

If AI is going to be good for anything, it must be about the production of data that helps the government manage the problems that we face as a consequence of AI. Otherwise, what is the point of having it?  We must use AI to empower the state to intervene to ensure that we get better outcomes; otherwise, we are failing everyone.

And what this means politically is that we also need a changed mindset amongst our politicians. The reaction to pressure must not in the future be an increase in austerity, and insecurity, and wage suppression. It must instead be, that we will invest to protect our future.

This is what is essential, and it's particularly essential with regard to our politics of care because the point is, AI could eliminate drudgery from the workplace, and if it does do that, then what we need to replace that drudge work with is work that is genuinely fulfilling. That would be a political choice and not a technical one, because the means to do this would exist in this situation.

So the test for government will become, will it treat AI as a reason to cut, or will it treat AI as a reason to promote resilience? A courageous state will create employment, productive capacity, care, and democratic accountability, using AI to assist the process; otherwise, we'll be in deep trouble.

AI, as I said at the outset, could create severe economic harm for the UK and other countries around the world.

It could create interest rate policy that would have untold consequences in terms of a recessionary impact.

It could increase profits for a few at cost to many, and so we require governments that will rise to this challenge.

They must respond to AI with an investment-led-anti-inflation policy that creates the new jobs that we need because people will want them to manage the consequences of AI having displaced their previous activity.

This is about justice as well as economics, and that therefore requires us to take action.

We must say that we will not accept an austerity narrative as a consequence of AI. That is unacceptable.

We must demand policies that sustain jobs and stabilise prices.

We must demand that capacity be built, whether that be in energy or homes, transport skills and care, and vitally, you can take part in all of this; that's why we make these videos.

You can now begin to answer the question:  Do we want to live in an economy fit for the AI age?

Share this video,   talk about it. Literally subscribe to this channel because there are many more things like this to come, and comment on YouTube and take part in the poll that's down below. Let us know your views. They matter, and that's the key point here. This is about you, and that's why this is important.

Poll

 Loading ...

Tickets are now on sale for the Funding the Future live event in Cambridge on 28 February. Tickets and details are available here. 

PDF of article