Economic transformation and the risks of loss of purpose

What will the nature of work become after a full economic transformation towards sustainability and a transition to AI?

It’s well-established that to avoid climate change impacts that will cause a social and economic tragedy, we must bring about a low-carbon transition. At the same time, governments around the world are considering how to embed artificial intelligence across their economies. The transition to Artificial Intelligence (AI) is mostly just very profitable and once the genie is out of the bottle, it’s hard to put it back in. But how are these transitions being designed? Are we prioritising transitions that make people to be happier, and create more fulfilling lives? Will inequality and the power of capital owners increase or decrease?

The reality is that there are different types of low-carbon and AI transitions that we can design and achieve. We have agency on how this plays out. But the question that we really want to drive home is, are we just a consumption society in the ‘West’, or in other words, is maintaining consumption the only thing that matters in those debates? What keeps us up at night is whether badly designed transitions could be damaging people’s lives with industrial decline, left behind communities, loss of hope, meaning and purpose. This, as we all know, leads to extreme politics. Politics of despair. This is an open debate, so let’s explain what we, as researchers, see happening.

The nature of capital-intensive technology transitions

First, let’s define AI and low-carbon transitions. Our understanding of AI is of a technology that organises large amounts of data and which can be queried in a variety of ways, which improves the ability of workers to carry out particular tasks (or which eliminates the need for some types of occupations altogether), generating savings or profits. AI is a natural extension of the digitalisation of the economy, but has a particularly important emergent feature of power concentration: it is almost exclusively only a few US and Chinese companies that provide AI services to the rest of the world. Low-carbon transitions for their part refer to replacing high greenhouse gas-emitting technologies and/or practices by low-emitting ones, in industry, government and households.

Low-carbon and AI transitions are both, not exclusively, but to quite a large degree, a story of rapid technological and social change. We will transform how we live, make things and use energy, almost entirely, over the next 20 years. This will happen very quickly. If we don’t actively transition economies, we get more severe climate change, extreme floods, heat waves, and other disasters, while we nonetheless still see rapid technological change, because it’s already committed and happening. We don’t know whether we will limit the worst of climate change, but a lot of low-carbon technological change is too far underway to stop by now. This can be seen in the transitions to electric vehicles, solar energy, wind energy, LEDs, heat pumps, hydrogen, and so on. AI transitions are also highly visible, and in fact, the two transitions begin to interact and mix through the emergence of intelligent low-carbon technologies such as self-driving vehicles.

Low-carbon transitions are about ‘getting rid of combustion’, and are rapidly underway. Therefore, under such scenarios, global fossil fuel demand necessarily peaks and starts to decline, and so would the fossil jobs. We also see other changes committed, such as the emergence of the internet of things (making objects communicate with one another), big data (where we record data about everything), industry 4.0 (or intelligent manufacturing) and AI more generally. These AI transitions are affecting other kinds of jobs, many in the services. If we adopt the terminology from the recent book by JF Mercure & Hector Pollitt, we can call all this together the ‘smartification’ of industry (‘smart’ as in smart devices) – making everything intelligent. This is broadly profitable largely because it saves input costs in industry (notably energy, materials and labour).

By the same token, smartification will kill a lot of jobs as it cuts industrial costs worldwide. To economists, of course, this is productivity growth. In the case of AI, this leads to a few large companies making unprecedented profits at the expense of a lot of wages saved, as workers are replaced while offering the same products. Say you’re a company manager. You can use AI to save the wages of a dozen employees. This saves you some money, which you keep as profit. Some of this goes to the AI provider, which they also keep as profit.

Many people’s wages are turned into two companies’ profit. What do these companies do with the profit? That’s the key question! They probably won’t be spending it on what the former employees were going to consume. The company could invest it in the construction of new real assets, in which case it is spent on wages again, creating jobs (e.g., construction workers). But this is often not what happens: there is a good chance it just buys paper assets in financial markets, which allows others to buy even more paper assets, and so on. This doesn’t create jobs. It stands in stark contrast to the cars, phones, food, restaurant meals or holidays that the laid-off employees might have otherwise purchased in the real economy.

Something similar can happen with the low-carbon transition. When consumers buy electric vehicles, the value loss in petrol or diesel that these vehicles will never consume means the equivalent amount of oil will never be drilled, so the wage of oil workers will never be paid. When fossil fuel vehicles and EVs cost the same (we are getting there now), someone else will keep that wage, as either profit or savings. It could be the consumer (lower expenditure on fuel, because EVs use a lot less electricity than petrol cars use petrol). It could, however, be someone else like the charging point operator. In the UK, for example, at 90p/kWh for rapid EV charge, the cost per km driven is roughly the same as petrol, but electricity costs closer to 25p/kWh, therefore a huge profit is made. If it’s the consumer that benefits from the saving, this will almost certainly get re-spent in the real economy and pay someone else’s wage. If it’s the charge point operator, chances are it will go into paper assets in the financial sector.

This is a ‘capital-intensive’ transition. We replace present and future wages by upfront capital (replacing people by machines). By definition, given that capital is owned by someone but workers are not, a capital-intensive transition concentrates revenue flows to capital owners, therefore increasing wealth and income inequality. For instance, the wage of the oil worker goes to trillionnaire Elon Musk via Tesla, or to Chinese shareholders via BYD. There is also a geographical dimension to this, as capital owners may often be concentrated in a few countries (the US and China for instance). A rise in the capital income share may also happen via prices and inflation (there is a range of mechanisms with which labour income can get converted into capital income).

This is not to say that the low-carbon or AI transitions should not be undertaken. Climate change needs solving and, as we have already established, a transition is inevitable in many sectors. But there may be a disaster around the corner with technological change. The current trajectory is not a rosy story of achieving sustainable development goals and climate change targets. It’s one about concentrating wealth. What happens when wealth concentrates and old outdated industries go in decline? We get the North of England and the Glasgow effect all over again, in the towns and cities where the jobs at risk are concentrated. This is a repeat of the Thatcher era for Britain: entire stranded towns with excess workers with outdated skills that they can’t apply anywhere, with nothing to do. It is really worth reading Shuggie Bain and Young Mungo, by Douglas Stuart, to see what deindustrialisation in Glasgow may have felt like in the 80s. It leads to loss of purpose, despair, substance abuse, while the towns, starved of investment, become run down. This is a very familiar story for many towns and cities in the UK. In the US too, cities in the Rust Belt, notably Detroit, suffered decades of decline, high crime rates, and deprivation following the loss of industry toward the end of the 20^th century. The loss of these jobs goes beyond the income that is lost. Communities and identities that have been built around work over decades get destroyed, fueling despair. Those impacted may then engage in extreme or populist politics because they feel nobody cares that they’ve been left behind. In the book by Mercure and Pollitt, this is explained with many important sociology references and a parallel is made between what’s coming and what happened to the coal and steel towns of Britain.

Where the economic analysis of technology transitions goes wrong

In much of the discussions of climate policy, comparisons are made between now and a low-carbon world, to see in which people would be better off. Would we be richer or poorer? These comparisons are usually based on GDP or consumption. But this assumes, in utilitarian spirit, that what animates people is only consumption. More consumption of trinkets makes people happy, less consumption of trinkets makes people unhappy (where the trinkets are imported from China). If you read any literature on economic anthropology or sociology, it rapidly becomes clear that this is not a very clever way to measure purpose and fulfilment. Are we truly only just interested in consuming? What’s the meaning of our lives’ achievements and daily work? Just working in order to consume trinkets, otherwise we would sit and do nothing? Why else would people dedicate their lives to caring for those closest to them, volunteering in their local communities, or expanding public knowledge for complete strangers?

Let’s do a thought experiment. Say we could be sure that after these technology transitions, we could consume a lot more trinkets, measured via their value. However, we’d all have had to change careers, start being paid to do nothing, or to do something really boring. How would that sound? Economists might assume that (1) this would be good, and (2) if there is more consumption, there will be more jobs, and people will be happy. We’re not sure that’s true.

Looking at purpose in work and the loss of skills

As mentioned, many towns and cities in the UK, Europe, and US suffered from deindustrialisation during the end of the 20^th century. Britain used to manufacture machines, leading the world in the late 19th century, and declining since the 1950s. If you find in your shed some old woodworking tools that say, ‘Made in England’, they are likely about 50 years old or more. Germany, until recently, also manufactured lots of tools, but increasingly, they are made in China, even if sold under the names of usual German brands. ‘We can’t compete on cost with China’, the companies will say. So now you either buy Chinese goods directly, or indirectly via Western brands. It is quite difficult to find anything manufactured in the UK, Europe or the US (in the case of the US, we mean other than software and AI).

This means that the UK factory workers have had to change careers from making tools, to selling tools. Economic theory will say, that’s ok, because the workers can continue consuming the same amounts of trinkets. In reality, that’s sad, because it means that much of the skills are disappearing. The cultures of manufacturing are disappearing. It’s irreversible, all this experience will be gone. It’s sad to say, if banking is the only skilled job one can have, while the rest is only about marketing whatever the Chinese make, it might become hard to find meaning in life (we’re exaggerating of course).

Economics says this doesn’t matter, because people continue to consume trinkets. Regional re-development is about allowing people in the deindustrialised provinces to consume more trinkets. But of course, happiness from trinkets is very much relative. If you become convinced that your purpose in life is to consume trinkets, because there is not much else to do, then this may become self-fulfilling. You begin to ask for your fair share of trinkets, because you’ve given up on the more meaningful things. For example, say someone wants to be a woodworker. There is no market for woodworking, because it simply cannot compete with dirt cheap furniture imports. Then woodworking and carpentry becomes an expensive hobby (consumption), rather than an occupation (output, investment). Some people were paid for it, now they must pay to do it.

The future of work and purpose in the West is bleak under current plans

Now, what’s the connection between this latter problem and the first, between jobs moving to Asia, and wealth concentration due to technology transitions? Well, it’s a bit the same thing. Jobs disappear, replaced either by nothing, or by possibly jobs, (say, manual workers redeployed to selling tools instead of making them), while GDP growth remains the same. Is this some kind of contradiction? No, because the concentration of capital income means that while the same total income flows in the country, it all goes to the few instead of the many. The economic models get fooled, but not people.

Now think of all the money going into AI. $86bn has been raised to build data centres in space (and colonise Mars as a side project). Where does all that money come from, in such a cash starved economy? The money is from ‘expected future wages saved’, or in other words, the wages of people we plan to sack. The future wages we intend not to pay to employees have already been paid in advance to the big AI owners, in what looks like religious belief in a dystopian economic model. But when everyone else’s lives lose purpose, the situation becomes ripe for the far right. But the US AI owners also happen to be involved in far right movements. So, it’s a good question where this is leading.

We are not just a consumption society. A consumption society would be where citizens are perceived as nothing more than consumers, while business and by policy-makers make decisions about people’s lives and purpose without much consent or accountability. The loss of demand for skills and occupations in the UK and Europe ought to be something that we think very carefully about, as we design technology transition. Generalised loss of purpose could lead to political chaos, and will backfire by either stalling emissions reductions, cause financial chaos, turn us to far right politics, or simply just make everyone miserable.

The solution?

We need to design technology transitions that do not concentrate wealth, but instead, transfer it to new meaningful activities. Given that by definition of those transitions, some occupations are bound to disappear (e.g. the oil jobs, easily automatable jobs), we urgently have to create new creative ones. They should be well planned industrial transitions, not least cost or max profit decarbonisation and diffusion of AI. We ought to think very carefully about what the occupations of the future will be, and how we want to use AI. We need to plan what oil drilling and car manufacturing workers will be doing in the future, and whether killing so many wages for the sake of ‘productivity growth’ is really wise, or whether it is going to backfire. Loss of purpose in working communities is something to consider carefully.

In economics, we promote converting labour income into capital income because we think this is productivity growth, and that productivity growth means more jobs. But this would only be true if wealth didn’t concentrate, and if profits were spent 100% by capital owners back in the real economy. When wages disappear and become spent on paper assets (or crypto assets), the purchasing power may never come out again into the real economy, as it remains trapped in the ownership of those already vastly wealthy. Instead, it inflates financial asset prices and circulates from one wealthy owner to another. For example, it makes houses unaffordable and therefore uninhabitable, as they gradually become financial assets rather than habitations. Meanwhile, loss of labour income could slow down spending and therefore GDP, as the wealthy do not consume nearly as much in the real economy as the wages of workers. The result is an increasing disconnect between the ultra-rich, who see everyone as only consumers of their products, and the rest, who may not want to be ‘just consumers’ and have a go at helping others and shaping tomorrow.