Will AI secure or shake Swiss prosperity?

AI could help ageing economies keep growing – or fuel mass unemployment and redraw the global wealth map. Which future awaits? Nasa/JPL

Artificial intelligence (AI) could close the output gap caused by an ageing population or create mass unemployment and turn the global distribution of wealth on its head. Which scenario will come to pass, and what will determine the future?

Behind every political debate in Switzerland that even remotely concerns the economy, the same unspoken feeling seems to emerge: the fear of losing everything.

Switzerland, a poor agricultural nation just over a century ago, lives with the paranoia of a gambler. Today, measured by both average income and assets, it is the richest country in the world.

Admittedly, wealth is unevenly distributed, and yet one in four pensioner households in Switzerland has over CHF1 million ($1.27 million) in assets. In most cases, this wealth is tied up in their homes. This “concrete gold” is largely due to having the highest mortgage debt in the world, with private households and banks taking a risky gamble on the future.

Even in the optimistic scenario, there are losers

So how stable is the economic future of a country with no natural resources and negligible international importance, at a time when AI and geopolitical conflicts are reshaping the world?

With this question in mind, we met with Jan-Egbert Sturm in Zurich. The director of the federal technology institute ETH Zurich’s KOF Swiss Economic Institute is regarded as one of the country’s most prominent economic analysts.

Sturm identifies four major risks for Switzerland – many of them shared by other successful economies: an ageing population, climate change, deglobalisation and artificial intelligence. Of these, however, AI is the only one he also views as a potential opportunity.

“The productivity gains from artificial intelligence could mitigate the negative effects of the other three trends,” he says. In recent decades, Switzerland has closed its productivity gaps through skilled immigration. “That has worked extremely well,” says Sturm. With AI, he believes the opportunities are now expanding.

>>Read this article about how demographic change and workforce migration are transforming Europe and Switzerland.

A recent study by consultancy firm Deloitte concluded that Switzerland could face a shortfall of around 300,000 workers by 2050. To compensate, labour productivity would need to grow by 1.2% annually, four times as fast as it has over the last 25 years. According to Deloitte, one way to achieve this is through AI.

But Sturm does not shy away from acknowledging that the AI revolution could also have negative effects. Certain occupation groups in Switzerland are already seeing rising unemployment figures.

Jan-Egbert Sturm has been director of the KOF Institute at ETH Zurich since 2005. zVg

“Five years ago, the advice was: ‘Learn to code, and you’ll be set for life!’ Today, AI can code faster and often more accurately than many IT specialists,” says Sturm, adding that demand for this skill is clearly falling. “We’ve been seeing this for some time now among interpreters.”

AI affects all skill sets, he says: “The question is: are you someone who will be more productive thanks to AI, or do you perform tasks that AI simply does better?” Sturm predicts a structural shift that will be painful for some, “but I don’t believe it will drive the unemployment rate up massively in the long term”. Even during the first IT boom 20 years ago, he says, it took a long time for the transformation to be reflected in the hard productivity figures.

Sturm therefore takes an optimistic view, much like Jensen Huang, the CEO of chip manufacturer Nvidia, who sees the idea that AI could be used primarily to make people redundant as a “lack of imagination”. AI automates tasks, but not professions, and companies should use the space freed up by this unburdening to pursue higher ambitions.

The World Economic Forum (WEF) also sees prosperity rather than an economic apocalypse on the horizon. In its Jobs ReportExternal link, the WEF predicts that 92 million jobs worldwide will disappear by 2030 due to macro trends such as AI. However, this would be offset by the 170 million jobs expected to be created.

The fine line between utopia and dystopia

However, there are also very different predictions. US tech investor Vinod Khosla made global headlines in spring when he claimed that a child who is five years old or younger today will never have to work in their lifetime. Was that an exaggeration designed to boost share prices, or a serious prediction?

We spoke to Peter G. Kirchschläger, an ethicist at the University of Lucerne who has spent years researching AI, and whose voice carries international heft. “The aim of artificial intelligence is not to complement but to replace humans,” he says, adding that we must recognise there will be a massive reduction in paid professional roles.

He says the situation is different from previous pivotal moments such as the Industrial Revolution, because for the first time, all functions and sectors are affected. Kirchschläger is certain that “dispensing with humans in certain areas will come with a loss of quality”. The question, he says, is whether AI solutions are simply so cost-effective that this is something we are willing to accept: “For me, that is a realistic scenario.” That would mean no more work for 50-70% of people.

Peter G. Kirchschläger is head of the Institute for Social Ethics at the University of Lucerne. zVg

Kirchschläger is not a naysayer. He advocates for a major societal transformation: “We must not now succumb to fear, but act rationally. The problem only arises if we do not take the change seriously and are not prepared for it.” If it comes to a large-scale loss of jobs, profits will be distributed even more unequally than they are today. New redistribution mechanisms would be needed, otherwise consumption would collapse and social unrest would break out. 

Kirchschläger proposes taxing data and AI and converting the revenue into a universal basic income on a global scale. It is a scenario in which wealth disparities would shrink, and with them the pressure to migrate. A new world order in which only a few – such as researchers and entrepreneurs – would have a job.

Because paid employment and income are decoupled, his model also includes an obligation to engage in community work, intended to preserve social cohesion and a sense of purpose. Finally, he advocates regulating AI globally based on human rights and establishing a UN agency to review and approve AI tools prior to their introduction, like the process currently in place for medicines.

It is a huge proposition when multilateralism is deteriorating before our eyes. Nonetheless, Kirchschläger remains optimistic. The pressure on states and the threat of social unrest, he argues, is simply too great. He points out that an agreement was also reached on nuclear power: “We managed to come to an agreement and establish global regulations. It’s not perfect, but the worst was avoided.”

The concept of a tactile renaissance

But which scenario will come to pass? Will there soon be no more jobs, or will AI simply boost our productivity? Can we draw any conclusions yet from current practice?

Marc Beierschoder is the AI and data lead partner for management consultancy Deloitte in Switzerland. He paints a picture of an evolution that is still in its infancy. His clients are focused on the here and now, he says: “Many have an innovation budget and want to explore how they can accelerate their digital transformation.” Others have specific cost-saving targets. For instance, one CEO demanded that AI be used to cut costs by 20%, he notes.

Marc Beierschoder is a partner specialising in AI and data at Deloitte Switzerland. zVg

AI is very powerful in certain fields, says Beierschoder, for example, in research or data analysis, and particularly anything to do with languages and dissemination of readily available information, such as internal training. Today, a technician who is stuck on a problem can tap into a company’s collective knowledge, or oversee a securities portfolio. For Beierschoder, it is clear that certain functions will disappear or become less common.

However, he believes that this growth opens new fields and creates new jobs. “We live in the world we have created. We could still be sitting in the trees today, eating fruit all day long. But no, we invented work. I don’t see a scenario where a superintelligence does everything for us,” he says.

The fact that one of his senior managers recently resigned to start an apprenticeship as a roofer – a job considered relatively safe from AI for the time being – does not contradict his assessment. He expects that repetitive and abstract functions, particularly in large companies, will be reduced, and that many people will turn to professions closer to their physical subsistence. “I am convinced that my daughter will work her whole life,” he concludes.

Widely diverging forecasts

Can this fear be dispelled? Hardly. When it comes to AI, there remains a tension between optimism and anxiety, as the Organisation for Economic Co-operation and Development (OECD), for example, notes in its latest Employment OutlookExternal link. Even the most positive forecasts come with the recognition that the value of human capital is declining. Knowledge painstakingly acquired over many years may suddenly become available for nothing.

If highly paid intellectual work loses value, the foundations of Western economies would be shaken –foundations that in Switzerland are propped up by millions in assets and vast levels of mortgage debt.

Adapted from German by Katherine Price/ds.