
Is Nvidia boss Jensen Huang right about a looming labour shortfall?
Loading summary
Tim Harford
This BBC podcast is supported by ads outside the uk.
Home Care Job Advertiser
You can make a difference in someone's life, including your own, with a job in home care. These jobs offer flexible schedules, health care, retirement options and free training. They also provide paid time off and opportunities for overtime. The Visit OregonHomeCareJobs.com to learn more and apply. That's OregonHomeCareJobs.com.
GoDaddy Advertiser
Running a business online look legit and own your own brand with professional tools from GoDaddy instantly build trust with your customers and boost your credibility with an email that matches your domain so people know you mean business. There's never been a better time. Just go to GoDaddy.com GDnow and choose from a wide variety of popular domains to find one that's right for you. Pair that with a professional email that works for all your business needs, from daily communications to email marketing and everything in between. That's a little price for a lot of credibility. For a limited time, get a domain and matching professional email for just 99 cents a month for one year. Go to GoDaddy.comGdNow and look legit with GoDaddy. That's GoDaddy.comGdNow again GoDaddy.comGdNow there's never been a better time to choose the domain and email that's right for you. New customer purchases only products auto renew separately. See terms on site godaddy.com gdnow.
Rajiv Gupta
Hello.
Tim Harford
And thanks for downloading the more or less podcast with a program that looks at the numbers in the news, in life, and in your jobs. I'm Tim Harford. Loyal listener Robert Gibbs emailed in to ask about a stat he heard in a talk by Jensen Huang, the CEO of Nvidia. Nvidia is a US Tech firm that's already become the most valuable company on earth by selling advanced chips to the companies developing generative artificial intelligence. Huang, its boss, was speaking at a conference in California where he told the.
Rajiv Gupta
Audience, we know very clearly that the world has severe shortage of human laborers, human workers. By the end of this decade, the world is going to be at least 50 million workers short.
Tim Harford
Huang was making the case for the futuristic solution to this problem, humanoid robots. But is the claim correct? Is the world really going to be 50 million workers short by 2030? Eager to justify our own jobs in journalism, we asked Nvidia where Hwang got this figure from, and they didn't reply, possibly due to vacancies in their public relations department. But we have tracked down a figure that looks remarkably similar in News articles from India citing a report by Boston Consulting Group, also known as bcg, a massive global management consulting firm. The report hasn't been made public, but its author, Rajiv Gupta, agreed to talk to us. So did he recognize the figures the Nvidia boss has been using?
Rajiv Gupta
We do not know if he's using our report, but yes, the number he uses does match the number that we have in our report.
Tim Harford
Okay, let's get into it. How did Rajiv work out that there would be a shortage of 50 million workers by 2030?
Rajiv Gupta
Like most estimations, it is a gap between what we project as the demand for jobs and the supply for jobs.
Tim Harford
Simple. First up, estimate the demand for jobs.
Rajiv Gupta
On the demand side. What we have done is we have taken the current employment data accurate as of 2024, plus all the open job postings.
Tim Harford
Using data from sites such as LinkedIn, Rajeev and his team worked out how many unfilled job vacancies there were in the whole world as of 2024. He reckons there are currently 10 million. Now, this figure is obviously a little ballparky already as not all jobs are advertised this way.
Rajiv Gupta
We then forecasted the GDP growth of different countries, adjusted it for labor productivity, and adjusted it for any impact of AI and technology to come out with an estimated employment growth rate.
Tim Harford
Not so simple. These extrapolations hinge on predictions about the future of AI. But no one really knows what's going to happen. With some tech soothsayers predicting unimaginable transformations over the next five years and some others suggesting it won't make much difference. Rajiv's analysis takes a middle way. He decided to treat AI like any other tech disruption, like the Internet, for example. And he went through different types of jobs, projecting the likely disruption for each.
Rajiv Gupta
Area and in some kind of industry. Across skills, the impact is as high as 50ish percent. In some, it is as low as 2 to 3%. So based on which industry, what type of work, we have estimated what kind of impact that can have.
Tim Harford
Some types of work are going to be hit hard by AI, such as software development and testing, some not so much, such as construction. And there'll be some new types of jobs too, for example AI prompt engineers.
Rajiv Gupta
So overall we have a point of view that the worker skills will change dramatically because of AI, but net it's going to create many more new jobs than it's necessarily going to take away.
Tim Harford
Okay, back to the 50 million. That's the first step. Done a necessarily massively simplified analysis of five years of economic growth and disruption. That gives you one plausible projection of the likely employment demand in 2030. Then you have to work out how many workers there'll actually be.
Rajiv Gupta
We took the labor flow rate data so different countries have different labor flow for women, for people of different ages and so on. And we adjusted it for both retirement and new entrants. To calculate the supply side of labor based on skills, they looked at the.
Tim Harford
Demographics of all the countries in the world and and the likely flow of new workers entering the system and older workers leaving it. Some countries have aging populations and declining workforce trends, some don't. They worked out three scenarios, conservative to aggressive, and came out with three estimates.
Rajiv Gupta
The numbers were close to 30 million, 40 million, 50 million.
Tim Harford
But what does Rajiv actually mean by a shortage of workers? Is he predicting an end to unemployment? That everybody will have a job and the world will still need 30 to 50 million more workers?
Rajiv Gupta
Well, no, we are not going down and looking at what will it do to the unemployment rate. We are just looking at the skills for open roles that have not been felt.
Tim Harford
So there could indeed be very high unemployment in some places and in some countries alongside a worker shortage. And this is where AI comes in again, because the worker shortfall is biggest in areas where AI is going to find it hardest to reach.
Rajiv Gupta
60% of the shortages comes from what we called blue collar industrial workers. So manufacturing, mining, cleaning, private security, then in sales professionals, retail store managers and so on. Then healthcare and education was another 12 to 15%, office administration was 10% and white collar workers, technology and so on was another 10%.
Tim Harford
And why are these sectors going to be short of workers?
Rajiv Gupta
It is the preference of workers in different countries and what they want to go into. So while we talk about labor shortage, there are areas white collar, highly skilled being one where there is likely to be a surplus of labor as well. But it is driven by the growth in some of these industries, let's say construction, manufacturing and so on, not being filled by individuals who necessarily want to get into that area.
Tim Harford
And to be clear, the worker shortfall isn't everywhere on the planet. It's concentrated in specific rich economies.
Rajiv Gupta
So 90% of the labor shortfall comes from 20 countries out of the 200 plus. Obviously, just given the size of the economy, USA is one large such country after us. UK, Germany, South Korea, Russia are the next and Poland, Hungary, Czech Republic put together. These are the next four or five countries regions and then followed by Japan, Australia, Canada.
Tim Harford
The bottom line is that Rajiv obviously doesn't have a crystal ball that's telling him how the global economy is going to shift and change over the coming years, but he was able to roughly trace out what happens if take the current trends and spool them forward a few years under a set of assumptions.
Rajiv Gupta
See, most estimates and extrapolations do turn out to be wrong, and I'd be humble enough to accept that. However, what this provides is a framework to think in what industry, in which country, what skills are likely to have a gap so that organizations and countries and workers can start thinking about it and maybe geared up towards it.
Tim Harford
One part of that might be that people are paid more to do the jobs not enough people want to do, although Nvidia boss Jensen Huang thinks a pay rise won't solve the problem and so we're going to need robots to do those jobs. But there are other options too. Higher levels of immigration could be used to fill these jobs, but that comes with all sorts of practical and political challenges.
Rajiv Gupta
A forecast of 45 to 50 million cap in labor is assuming no significant interventions, but actually the report then goes on to recommend interventions. So I do hope and pray that by 2030 the number is proven wrong and those interventions have actually happened.
Tim Harford
Our thanks to Rajiv Gupta from Boston Consulting Group and that's all we have time for this week, but if you see any suspicious stats, please let us know at more or lessbc.co.uk. we will be back next week and until then, goodbye.
Grainger Advertiser
This is the story of the One as head of maintenance at a concert hall, he knows the show must always go on. That's why he works behind the scenes, ensuring every light is working, the H Vac is humming, and his facility shines with Grainger's supplies and solutions for every challenge he faces. Plus 24. 7 customer support. His venue never misses a beat. Call quickgranger.com or just stop by Granger for the ones who get it done.
Home Care Job Advertiser
This is the story of the 1. As a custodial supervisor at a high school, he knows that during cold and flu season, germs spread fast. It's why he partners with Granger to stay focused, fully stocked on the products and supplies he needs, from tissues to disinfectants to floor scrubbers, all so that he can help students, staff and teachers stay healthy and focused. Call 1-800-GRAINGER click granger.com or just stop by Grainger for the ones who get it done.
Host: Tim Harford, BBC Radio 4
Guest: Rajiv Gupta, Boston Consulting Group
Date: September 13, 2025
This episode investigates a striking claim made by Nvidia CEO Jensen Huang: that the world will be short of 50 million workers by 2030. Tim Harford and the More or Less team examine the statistic’s origins, its credibility, the calculations behind it, and what such a shortage actually means in a rapidly changing world influenced by the advance of artificial intelligence (AI).
A. Estimating Demand for Workers
B. Estimating Labor Supply
The ‘shortage’ doesn’t mean that unemployment will disappear; rather, it points to a gap in available skills for specific open roles.
High joblessness and a worker shortage can co-exist, especially as the deficit is concentrated in roles that are hard to automate:
The paradox: some sectors will have excess labor (especially high-skilled/white-collar), while manual/industrial roles remain unfilled.
The shortfall is concentrated:
Rajiv Gupta acknowledges the limitations (“most extrapolations do turn out to be wrong”) but emphasizes the value of the framework for workforce planning.
Policy and market responses could close or reduce the gap:
The reported forecast assumes “no significant interventions”; active measures might substantially lower the estimated shortfall.
The headline claim of a 50 million global worker shortfall by 2030 is a plausible projection based on certain trends and assumptions, though by no means inevitable or precise. The gap mainly affects specific roles and countries, and is influenced by demographics, economic growth, worker preferences, and the unpredictable impact of AI. Policy choices, market responses, and technological advances between now and 2030 could dramatically reshape this picture. The episode encourages skepticism toward big numbers in headlines, and offers a useful model for thinking about workforce change.