top of page

UX Roundup: Wealth and Happiness | Wealth and AI | ASI and Unemployment

Writer's picture: Jakob NielsenJakob Nielsen
Summary: The relationship between wealth and happiness | Wealth and AI | Superintelligent AI and human unemployment

UX Roundup for January 27, 2025. Happy Year of the Snake! May the new year bring you happiness and prosperity — the two are usually linked, as my first news item shows. (Midjourney)


Listen to my Year of the Snake UX Song (YouTube, 2 min.)


Wealth and Happiness

There’s an old saying that money can’t buy you happiness, but that’s wrong. At least in the aggregate, more wealthy people tend to be happier. The website “Our World in Data” contains a profusion of interesting information about many subjects, including the relationship between money and happiness.


I’ll discuss one chart, which I replotted based on data from 133 countries in 2022 presented in the linked article, which again sourced its data from the World Happiness Report and the World Bank.

In this chart, each dot represents one country. The x-axis is the average wealth in each country, measured as GDP per capita at purchasing-power parity (i.e., how much you can buy for the money in that country). These numbers are plotted on a logarithmic scale. The y-axis is the average self-reported happiness score in each country, measured on a 1-10 happiness scale, with higher numbers indicating happier people. (Hat tip to Steven Pinker for alerting me to this data.)


As you can see, there’s a very strong correlation between wealth and happiness. The correlation between log(GDP/Capita) and Happiness is r=.83, which is astoundingly high. Simply knowing the average GDP in a country allows you to predict 69% of the variability in people’s happiness. (That’s the orange regression line on the above chart.) Of course, this leaves 31% of happiness to be caused by other factors. (The linked article discusses many smaller factors, such as national culture and political freedom.)


The key point in interpreting this chart is that money is a relative concept, in terms of relating to happiness. It’s not adding a certain amount of money that makes people happy. It’s adding a certain percentage. That’s why we must plot money on a logarithmic scale.

As an example, earning an extra $1,000 will make essentially no difference to a person who already makes $100,000 (typical in countries like Singapore, Switzerland, and the United States). Particularly after taxes, you can hardly feel that extra money and any extra happiness would be so minuscule that it wouldn’t show in the chart. On the other hand, a person making $1,000 per year (typical in countries like Congo, Malawi, and Niger) would get twice as rich from earning an extra $1,000. This would likely add about half a point to that person’s happiness on the 1-10 scale.


For an American to gain half a point’s worth of happiness, doubling income — or making an extra $100,000 in our example — would also be required.


What does this data have to do with user experience? One of the main outcomes of good UX design is to improve productivity, which is the only way countries get richer in the long run. Yes, some people are lucky to live on top of large deposits of oil or diamonds, but such riches are a one-time windfall that won’t get better next year. On the other hand, if every year employees produce 5% more for each hour worked, that accumulates, and only 15 years will be required to double income, which would, on average, make people half a point happier. Sustain these productivity improvements for 30 years (less time than the average career), and people have gained a full point of happiness.


Rich countries have sadly not been able to sustain 5% annual productivity gains in recent decades. However, with the shift to an AI-fueled economy, the high productivity gains of the past are likely to return.


The point that better design and higher productivity create happiness isn’t a short-term phenomenon. Even if everything goes well, you won’t feel notably happier next year. But in 30 years, you will likely have gained a full point of happiness if we all do our work right.


“Rich and happy” Midjourney’s interpretation of this prompt. Sometimes, I like to see what AI comes up with when I let it lose.


“Unimaginable Wealth” Is Easy to Imagine

Some of the more, lets say, “excitable” AI influencers have started posting about the unimaginable wealth we’ll get once AI moves up a step or two, and intelligence becomes almost free. For sure, we’ll get service as software, meaning that services that used to require expensive humans will be delivered by cheap AI. Anything we can dream up, whether software applications or movies, will be created in a matter of seconds.


Immense progress in science and medicine will happen overnight, once we can assign thousands of Ph.D.-level AI agents to work on any research question. And with robots, the cost of production will drop for any physical product.


Add all these improvements up, and we’ll gain a factor of 100x growth in our standard of living over the next 10-20 years. That’s the claim, though I personally think that a doubling of living standards over the next decade and maybe a 5x gain over the subsequent decade (for a total of 10x over 20 years) is more realistic, because companies and organizations change more slowly than the technological capabilities hypothetically could deliver if there were no breaks on progress.


Whether you believe AI will give us 10x or 100x in added wealth over the next 20 years, in any case, it’s a lot. I’ve seen this referred to as an unimaginable increase in wealth. However, it’s actually quite easy to imagine a 100x increase in wealth, because we can already see how this magnitude of wealth difference looks in real life.


Referring back to my first news item in this newsletter, there’s currently roughly a factor of 100x difference in the average income per person in the world’s richest countries like Singapore, Switzerland, and the United States, relative to the most dirt-poor countries like Congo, Malawi, and Niger.


The difference between medium-rich countries like the EU and Canada compared with medium-poor countries like India isn’t nearly as great, but it’s still a factor 6x, which is close to my personal prediction of AI causing a 10x improvement in living standards over the next 20 years.


Thus, the most enthusiastic take on AI’s potential for improving humanity’s existence corresponds to lifting the citizens of Congo to become as rich as Singaporeans. And my more modest prediction corresponds to lifting the citizens of India to become as rich as Europeans. Both are easily within the reach of credible imagination.


AI will generate so much wealth that it’s tempting to refer to it as “unprecedented,” if it were not for the fact that there are plenty of precedents for a 10x — or even a 100x — improvement in the standard of living, both across history and between countries in the world today. (Leonardo)


A second take is that of history. There are reasonable estimates of GDP per capita in the U.K., going back to the days when Britannica was a province of the Roman Empire. For the first 1,300 years of this 2,000-year period, people in the U.K. were approximately as rich as people are in the Congo today, clocking in at $1,000 GDP per capita in inflation-adjusted dollars at PPP.


Britain’s GDP per person started to increase slowly around year 1300, doubling from about $1,000 in year 1300 to about $2,000 around year 1665. Then the early Industrial Revolution started, and GDP grew by 20x to slightly less than $40,000 by 2019. (Remember that all these amounts are adjusted for inflation at purchasing power parity, or PPP.)


GDP per person in the United Kingdom (or the Roman province of Britannica) over the last 2,000 years. Plotted based on data from “Our World in Data.”


Thus, modern Britons are about 40 times as rich as their Roman-era or Medieval ancestors. Compare the U.K. now with the country back then, and that gives us another impression of the impact of large-scale economic growth.


The main difference is that we can now expect growth of maybe 10x in 20 years, as opposed to the growth of 20x the U.K. racked up over a 350-year period. Things move faster now.


Well-off people four hundred years ago (except for the King) had a pittance of wealth compared with people today. Similarly, rich countries today will soon be considered poor compared with the standard of living we’ll enjoy in 20 years if the AI revolution plays out fully. (Leonardo)


Superintelligent AI Won’t Make Humans Worthless

Around 2030 we expect to achieve “ASI” — artificial superintelligence — which means that AI will be smarter than any living human. Once we’re reduced to the status of comparative ants in the rank-ordering of intelligence, will humans be destined for the garbage heap of history?


By 2030, humans will be the equivalent of ants in the ranking of intelligence. What will that mean for us? (Leonardo)


Maxwell Tabarrok recently published a good article addressing the issue of humans versus smarter AI. He has a variety of arguments for why humans will still be needed and have jobs, but I found his historical analogy particularly compelling:


We’ve already seen what happened to humans when many higher-intelligence beings joined the workforce. This has happened to stupid humans in a major way during the last thousand years.


For simplicity, let’s define people with below-average intelligence (IQ<100) as “stupid” and people with above-average intelligence as “smart.”


With this simplistic definition, there are about 1.8 billion smart people in the workforce. With this many smart people available, how could any stupid person ever hope to have a job?

More to the point, there were only 0.3 billion smart people in the workforce back in 1800. Over two centuries, we’ve added about 1.5 billion smart workers. This huge inflow of “superintelligent” workers (relative to the stupid ones) should surely have driven every stupid worker into unemployment.


Of course, this didn’t happen. We have many more smart people in the world than we used to, and yet stupid people still have jobs. Unemployment continues to be low, decade after decade. (Even when unemployment is “high” and everybody curses the politicians, it still tends to be less than 10% and most of those unemployed people get jobs a few years later.)

Why do stupid people still have jobs when there are so many new smart workers available? Because we’re richer now than in the past and many new occupations have been invented that soak up all the smart folks. Yes, a high-IQ person could perform a low-skilled job, but there are many new and better jobs available that they prefer.


Even more important, not only do stupid people still have jobs, they have vastly better-paid jobs than in the past. Yes, high-skill jobs pay better than low-skill jobs, but the jobs typically held by somebody with IQ 90 allow that worker to enjoy an incredibly higher standard of living that an equivalent stupid person had in 1900. (Or any year in human history, prior to now. The good old days were not good unless you were the King.) The many inventions and improvements created by smart people are also available to stupid people, so their labor becomes more valuable and the things they buy become cheaper.


In the future, with superintelligent AI, a similar situation will occur. If the fixed-labor fallacy hadn’t been a fallacy, then those AIs would indeed take most jobs. But since the concept of a fixed amount of work needing to be done is indeed a fallacy, there will many new tasks performed in the economy, and humans will do enough of them to still be needed.


Think back to my prediction about the future of product design jobs: By 2040, AI will likely perform 75% of the tasks that used to be done by UX professionals. And yet I predict a 5x growth in UX employment, because I also predict that 20x more good design work will be bought by the richer world that will come into being because of AI. Being a luxury good — such as quality design — is beneficial when things keep getting better.


Every labor-saving invention has led to predictions of unemployment, as fewer employees would be needed to do the same work as before. However, every productivity enhancement has also created more wealth and new opportunities, leading to a net increase in jobs — even if not the old jobs of carrying rocks around by hand that were made obsolete by our inventive caveman. (Leonardo)

 

Top Past Articles
bottom of page