This article continues the discussion I started in my previous piece, The Consequences of the Mass Adoption of AI and Robotics for the Average Person, where I looked at how AI and automation could change the economic value of human labor. Here, I want to focus on the next question: if large numbers of people do lose stable, well-paid jobs, what could realistically replace the current model of mass employment?
When people talk about the mass adoption of artificial intelligence and robotics, economic romantics almost automatically come up with the same answer: if people lose their jobs, the government will start paying everyone a universal basic income. In other words, people would receive money simply for existing—with no strings attached and regardless of anything else.
Sounds logical? Of course. If machines produce goods and services while human labor becomes less and less necessary, why not simply redistribute part of the wealth they create among everyone? But is it really that simple? I seriously doubt it.
The problem with universal basic income is not only whether the state can find the money to pay for it. There are other, much more important questions: how much are we actually talking about, and what will ultimately back those payments? If the government pays everyone, say, $100–300 a month, that is perfectly feasible for a strong economy. But such an amount does not solve the problem of someone who has lost a decent job and can no longer afford housing, support a family, and cover everyday expenses. If, on the other hand, we are talking about a genuine basic income—an amount a person can actually live on—the scale of the cost becomes entirely different.
And this is where a paradox emerges: the poorer a country is, the harder it may be hit by automation, while at the same time being less capable of financing a full-fledged basic income.
Where Could Universal Basic Income Actually Work?
It seems most realistic in wealthy countries with high productivity, strong institutions, and substantial income from capital or natural resources. Especially in places where the state already owns large investment funds, infrastructure, or a significant share of national capital. In such a system, one can argue that part of the economy’s income belongs to all citizens, and therefore everyone is entitled to a share.
This is fundamentally different from a system in which the government simply raises taxes on those who still work in order to support those whom the economy no longer needs.
For large developed economies—the United States, Germany, France, the United Kingdom, Japan—a basic income is technically possible, but politically far more complicated.
Because the questions arise immediately: Why should an unemployed person and a millionaire receive the same payment? Should a basic income replace pensions, unemployment benefits, subsidies, and other social programs? How high would taxes have to be?
And perhaps most importantly: would society be willing to support a substantial part of the population for decades simply because there is no longer an economic need for them to work?
This is where I see the main problem. Universal basic income is often discussed as an economic mechanism. In reality, it is first and foremost a political and cultural question.
What If There Is No Basic Income?
In my view, this is an entirely realistic scenario. We are used to thinking that if technology radically transforms the economy, the state will inevitably respond by creating some new, coherent system. But history tends to work differently.
Governments rarely redesign the social order in advance. Usually, they react only after a problem has become widespread, painful, and politically dangerous. So it is entirely possible that there will be no smooth, carefully managed transition. There may never be a moment when a government announces: from now on, robots produce enough, so every citizen will receive a guaranteed income. More likely, the process will unfold in a much more chaotic way.
First, a person loses a good office job. Then they find a worse one, and their income falls. Perhaps they qualify for a housing subsidy, some form of child support, or temporary unemployment benefits. The government offers retraining courses. They pick up occasional work through an online platform. Over time, even that work becomes scarcer, leaving them to survive on irregular small jobs.
Formally, this person may not be unemployed. But their economic status will be completely different. And this is exactly what I see as one of the most likely scenarios of mass automation.
Does Automation Necessarily Mean Mass Unemployment?
We may never see unemployment reach 30 or 40 percent at all. Instead, what we may see is a mass deterioration in the quality of employment.
A former accountant may end up working in customer support; a former junior developer may be checking AI-generated output; a former designer may be taking small jobs on online platforms or, again, reviewing the results of AI generation; a former office worker may end up in delivery or other service jobs. The statistics will show that all of them are employed. But this is no longer the middle class that existed before.
So, in my view, the main threat of automation is not necessarily the complete disappearance of work. A far more realistic outcome is a massive decline in the value of human labor.
So Why Not Simply Shorten the Workweek?
This scenario seems far more realistic to me than the classic idea of universal basic income.
Imagine that, thanks to AI, one hundred people can do the work that previously required two hundred. What do you do? The simplest answer for a business is to fire half the workforce. But society could, at least in theory, choose a different path: instead of cutting the number of employees in half, it could reduce working hours. In other words, rather than having 50 people work 40 hours a week, 100 people could work 20–25 hours.
For the owner of a company, the first scenario is obviously more attractive. For society, the second is far more stable. That is why I would not be surprised if one consequence of automation were a gradual transition first to a 30-hour workweek and eventually to something considerably shorter.
After all, we have already gone through a similar transition before.
There was a time when a six-day workweek seemed normal. Then came the five-day week. And instead of working 12–15 hours a day, we ended up with 7–8. There is no natural law stating that a human being must work exactly forty hours a week.
What If a Person Still Does Not Earn Enough?
In that case, a more realistic alternative to UBI may be not a completely unconditional payment, but a guaranteed minimum income.
The logic is simple. A person works but earns too little—the state tops up their income. As their earnings rise, the benefit gradually decreases. This is much easier to justify politically than a system in which everyone receives the same amount regardless of income or employment status. It also preserves the link between work and income.
It seems to me that, politically, this model has a much better chance of actually being implemented.
Maybe People Would Not Need That Much Money in the First Place?
This is another important possibility. We often think about the problem of income only in terms of wages. But if a person has access to healthcare, education, transportation, the internet, basic housing, and some essential utilities, they need far less cash income to maintain a decent standard of living.
So instead of universal basic income, we could see the development of a model based on universal basic services. Not €2,000 in cash every month, but a much smaller payment combined with a system of services that people do not have to purchase at full market prices.
For Europe, this scenario seems far more realistic to me.
Can the State Simply Create Jobs?
Of course it can—history has seen this many times, including in our country’s recent past. We may have to return to this approach as well.
If the private economy needs fewer and fewer people, the state can become a kind of employer of last resort. And this does not necessarily mean pointless jobs created merely to improve employment statistics. There are many areas where the social value is obvious but the market return is relatively low: caring for the elderly, supporting children, education, environmental projects, community work, culture, public-space maintenance, sports, social assistance, and so on.
We may even have to reconsider what we mean by productive work in the first place. A person who spends eight hours moving numbers from one corporate spreadsheet to another is considered a productive worker because a company pays them. A person caring for an elderly mother, meanwhile, is formally producing nothing.
That is a rather strange definition of economic value.
Who Should Receive the Profits from Robots?
In my view, this will be the key question. We can argue endlessly about robot taxes, basic income, or subsidies, but if automated capital produces an ever-growing share of society’s wealth, the most sustainable solution would be to ensure that the population owns at least part of that capital—through sovereign wealth funds, pension funds, national investment funds, or public ownership of infrastructure. This could also include stakes in energy companies, data centers, and automated manufacturing.
The logic would then be not: “a robot took my job, and the state took money from its owner and gave it to me,” but rather: “a robot replaced part of my labor, but as a citizen I receive a share of the income generated by the capital that robot represents.”
And this is a fundamental difference. In the first model, society remains in constant conflict between the owners of capital and those who depend on redistribution. In the second, the population itself gradually becomes a co-owner of the automated economy.
What If None of This Happens?
This possibility also needs to be stated plainly. There may be no social dividend, no shorter workweek, and no effective system of guarantees.
Instead, we may end up with a society in which a relatively small group of owners, highly skilled professionals, and managers earn enormous incomes, while a large share of the population lives somewhere between low-paid work, social benefits, and irregular gig income.
In other words, people may not be starving. They will still have phones, cheap entertainment, basic goods, and some degree of social support. But the path into the middle class will become much narrower.
In my view, this is a far more realistic dystopia than some futuristic world in which billions of people are simply left with no income at all.
So What Could Replace Mass Employment?
I do not think the future will offer a single universal solution. More likely, we will end up with a mix of different mechanisms: a shorter workweek, income supplements for low earners, affordable basic services, public-sector employment in socially useful areas, and the gradual expansion of public ownership in parts of automated capital.
And if universal basic income does appear, it will more likely be one element of this system rather than its foundation.
Because the main problem of the future automated economy, in my view, is not whether we will be able to produce enough goods and services. Quite the opposite: we may be able to produce them on a scale never seen before.
The main question will be different:
on what basis will a person be entitled to a share of that wealth if their labor is no longer needed to create it?
The modern economy is built on a simple formula: a person sells their labor and, in return, gains the right to consume. If artificial intelligence and robotics begin to break that formula, we will have to come up with a new one.
And I suspect that this—rather than artificial intelligence itself—may become one of the defining political conflicts of the 21st century, bringing the left, the right, liberals, and centrists into the same arena.
And, as always, the dollar will win.
