A world without paid work

Summary

• Daniel Susskind’s latest book A World Without Work – Technology, Automation and How We Should Respond (Allen Lane, 2020) was published just before the pandemic. In his book Susskind argues that technological progress will inevitably lead to a situation where there is less work to go around.
• Dr Daniel Susskind is a Research Professor in Economics at King’s College London and a Senior Research Associate at the Institute for Ethics in AI at Oxford University. Previously, he has worked in various roles in the British Government, including as a policy advisor to prime ministers.
• In Tampere, Susskind will focus particularly on the impact of AI technologies on the labour market and education.

The central argument Daniel Susskind makes in his book A World Without Work is that breakthroughs in artificial intelligence mean that all kinds of jobs are increasingly at risk. Going forward, machines will be able to perform many tasks that until recently were thought to require human involvement. Computer programs can compose music and design buildings, diagnose diseases and identify birds mid-flight.

In short, automation no longer threatens to take over just routine tasks, but also tasks that require considerable creativity, judgment or even empathy.

Instead of predicting that the AI revolution will result in a massive loss of jobs, Susskind suggests that a gradual withering will occur across multiple sectors. In some sectors, jobs may gradually disappear altogether, while in others, such as care, it is far more diffiult to replace humans with automation. Regardless, it may be impossible to persuade people to switch careers from, say, manufacturing, to the care sector. Jobs might also be available in places where people don’t want to live. In other words, while there may be no shortage of work, there may be a labour market mismatch.

Since technological progress is creating widespread unemployment, Susskind believes that educational reforms are urgently required to respond to the changes in working life.

Given the erosion of many kinds of work which looks likely to happen with AI, do you think we will see a similar decline as we did in manufacturing and industrial jobs? What might this decline look like? And how can we prevent this from happening?

It is often said that work is not simply a source of income, but also a source of meaning and purpose in life. And if this is the case, then not only will automation deny people an income, but it may also leave some people without a sense of direction and a sense of fulfilment in life.

In fact, a big part of my book, A World Without Work, is about the fact that states should look at this issue of meaning and purpose and try to understand it better. One of the most provocative and speculative, but nevertheless important arguments of the book is that the state must take a greater interest in these issues.

In the 20th century, states developed a whole variety of labour market policies to shape the way people worked. They also developed labour market policies that shaped people’s leisure lives. And if artificial intelligence starts to replace people in the world of work, the state needs to take a bigger role in regulating people’s spare time, too. So, to accompany labour market policies, we will also need leisure policies.

If we accept the idea that some people might find themselves without work of any kind, then leisure is something we will need to think about in more detail.

If the current trajectory in the development of AI continues, will it inevitably create inequality?

Yes, I think it’s already creating inequality today. The argument I make in my work is that we need to take technological unemployment seriously, in part because of the extraordinary technological inequalities we already see today.

The labour market is the main way of distributing income in society. For most people, their job is their main, if not their only, source of income. The vast income inequalities we see around us today show that this approach is already creaking, with some people getting far more for their efforts than others. To some extent, technological unemployment is just a more extreme version of this issue. It’s a story of inequality that ends with some people receiving nothing at all.

Economist Daron Acemoglu suggests that we should make determined efforts to redirect the development of generative AI. In his opinion, the displacement of humans with artificial intelligence and automation is not inevitable; instead, he believes it is entirely possible to redirect AI development onto a human-complementary path. How do you see this? Is redirecting required and is it achievable?

I’m very sympathetic to that argument, particularly in the short and medium term. I’m not a technological determinist; I don’t think that technological progress has a certain predetermined direction. On the contrary — the direction can change, and we can influence it by changing the incentives available to computer scientists and business leaders.

Where I depart from Daron a little is whether in the longer run we can continue to redirect technological progress, or whether at some point these technologies are just going to become so extraordinarily capable that our attempts to redirect them in a pro-worker way will hit a dead end. But certainly in the short and medium term, I think redirection is a big issue.

In fact, my new book called Growth: A Reckoning, which will be out in 2024, is all about how we can redirect technological progress, not only to meet the challenge of disruption in the labour market, but also with respect to climate change, AI in politics, and the loss of thriving local communities and places. So I think there’s an interesting role for redirecting technological progress more generally.

For a long time, people thought that computer programs could not be creative, and that creativity would always remain something inherently human. However, what we saw as the first real breakthrough in the current wave of generative AI was image creation with tools such as Midjourney and Dall-E. The results are often truly surprising and at least seemingly creative. In your book A World Without Work you give many older examples of creative work done by machines. Considering you have researched this subject matter for a long time, have you been surprised by any of the recent developments in generative AI?

I’ve been working in this field for about 15 years, and until very recently most people maintained that machines will never be creative in the way humans are. And yet, what has been so striking about generative AI in general is how it can solve problems that we use creativity to solve. As we have seen, it is indeed capable of producing original text and images. I haven’t been too surprised by that, because I’ve been arguing for some time that this is coming.

One of the things I do find particularly impressive is its ability to be funny. Humour, the ability to tell a joke and get a joke, has been considered something deeply human. But surprisingly, these programs can be quite funny on command! I find that baffling.

But in general, these technologies are forcing us to engage in a lot of self-reflection about what it really means to be human. What is inescapably human and what is not.

While generative AI is capable of writing text that passes for a news article, publishers seem to broadly agree that they wouldn’t publish AI-generated news without human involvement. There are many good reasons for this, including the inability of AI to tell the difference between fact and fiction. But since it seems likely that readers can’t tell the difference between text written by a human or created by AI, will companies in the news business be tempted to publish something “good enough” as more content creates a higher click-through rate for display ads and thereby more revenue?

I don’t think it’s a future temptation. It’s a temptation that exists today and some large companies have given in to. If you go to a site like Bloomberg.com today for your financial news, about a third of the content that you read is now generated by an automated system. And I imagine many readers, exactly as you say, wouldn’t be able to tell the difference between that and text written by a human journalist.

Now, there are clearly some areas of journalism where automation is much harder and less likely. But at the same time, there are many areas where automation is indeed possible.

The development of AI is extremely fast-paced and somewhat unpredictable. What career advice would you give to a young person trying to figure out their professional path?

My children are five and a half and two and a half, so at the moment I’m more concerned with getting them to sleep than with their future careers. But it is something that’s on my mind from a personal point of view, too.

Broadly speaking I believe that the challenge in the next ten or twenty years is not the disappearance of jobs. The challenge is that there will be jobs available, but for various reasons people aren’t going to be able to do those jobs. Technology is going to create new jobs, but those are not for everyone. One of the major reasons is that people lack the right skills and capabilities.

I think the big thing for young people is to make sure that they are learning the skills and capabilities that are going to be valuable and important. Very crudely, I think there are two strategies. Either people can try and compete with these systems and machines to do the sorts of things that they cannot do, or learn how to build and operate these systems and machines.

Now, that distinction might sound a bit too straightforward, but if you look at what we do in education and in workplaces, you’ll see that we’re not focusing on either of these strategies. Often we encourage students to practise the routine activities that these technologies are already very good at doing. Consider, for example, what the first five years of being a junior lawyer looks like. Document retrieval, document assembly, document review. All of these are tasks that new technologies are already encroaching upon.

There is a big burden on all our educational institutions, from early childhood education to workplaces, to take the sorts of educational reforms that are required seriously. This will be a big theme in my remarks at the Tampere Conversations event.

Text by Johanna Vehkoo
Translation by Leni Vapaavuori and Nick Moon