The current state of the employment market is quite interesting. The data on junior employment, companies that are adopting AI and really drinking the Kool-Aid of AI is really only moving in one direction. And right now, across the S&P 500, headcount fell in 2025 for the first time since 2016. That's a net reduction of around 400,000 jobs, ending eight consecutive years of just constant workforce growth. This discussion is going to be one of the most important discussions to have as a society over the next few years. And there's some really compelling arguments against worrying about it. Some of the people that I'm going to mention in today's episode have discussed the fact that there's been many waves of technology that's reshaped, for example, programming, and often thinks the doomsday narrative misreads history. But both things can be true at the same time. There can be an optimistic and pessimistic view of the future. Today's episode is going to explore those views, the data points and the history behind them, and why the companies that are cutting junior headcount are going to make a decade-long mistake. This episode brings together many forms of the latest research and thinking in this field to try and present the most up-to-date view. This is In The Loop with Jack Halton. I hope you enjoy the show. So I guess let's start off with some of the context and the data behind where we're at today. And I think a really interesting place to start here is a fantastic comparison that I found online and I fell in love with it. So in January 1971, a billboard got put up on the, I was going to say motorway, but in the US they're called highways around Seattle. And it read, were the last person leaving Seattle turn off the lights? And, you know, Boeing, the context behind this is Boeing had started 1969 with 134,000 employees and ended in 1971 with about less than 80,000. So, you know, about 44,000, 45,000 employees have been let go in 18 months. They wound down the space program. They had a big supersonic, you know, transport project that had been completely cancelled. And Boeing claimed they didn't need engineers anymore. But you look at today and the aerospace industry is projecting a shortage of more than a million engineers by 2030. 25% of the current workforce is over 55 years old. and you know that the cohort that would now say be the senior experienced employees never ever got hired and that really sets the scene for today's discussion you know why not hiring junior employees is going to really hurt the industry in 10 15 or 20 years and this episode isn't about just whether or not ai is destroying jobs which is often the version of this i guess conversation that generates a lot of big headlines and noise. It's actually just talking about the realities of talent pipelines and specifically what happens when the entry point of a career is closed before people have had the opportunity to start on their journey. As I said, the S&P 500 across as a whole in headcount fell in 2025, which was the first since 2016, a net reduction of about 400,000 jobs, which is ending eight years of basically uninterrupted growth. And that's a massive signal. We've had multiple research projects released in the last year, all pointing broadly in the same direction, which is at companies that have adopted AI, junior employment dropped between about 7% to 10% within about a year and a half. While senior employment at those same places keeps rising It doesn take a genius to figure out that that isn sustainable And this decline by the way in most places has happened through slower hiring rather than just redundancies as a whole So people have stopped bringing people in at the bottom. And we've had another study, for example, that drilled into the specifics of how that's happening. And they found that workers aged 22 to 25 in an AI-exposed job, and we can discuss what that is defined as shortly, are basically down 13% relative to their peers in less exposed jobs since 2022, 2023. For example, 22 to 25-year-old, let's say, software engineers, it's around 20%. So that's like nearly 25% of the developers that would have been hired and no longer being hired anymore. That's a huge cohort of people. and a really good telltale here is that the gap between where employment should be on let's say pre-ChatGBT employment trends is 500,000 less jobs than was expected so there are 500,000 less developers than there was predicted based on the current trajectory before ChatGBT released and this is a huge mistake for the industry like yes maybe some job roles are going to be compressed and maybe there are new job roles that will emerge, but not hiring junior employees and training people is obviously unsustainable. So let's explore why this is happening to junior employees specifically, because it's less obvious than you'd think. Let's take Luis Garciano, who presented a really interesting argument here. And Garciano is a Spanish economist at the London School of Economics, and his main work has been exploring how organizations structure knowledge so who knows what who does what you know how information might flow through a group of people and he said a job isn't a list of tasks that can be chopped up and automated arbitrarily it's a bundle so a bundle of tasks that are entangled into each other the question ai poses isn't can this task be automated it's can this task be separated from the rest of the job without unraveling the whole bundle of tasks. And basically what he's saying here is that a job isn't just a set checklist. You know, some parts are completely independent and freestanding. So you can lift them out, hand them to someone else, and nothing like breaks down. Whereas some parts are woven into everything else and the relationships, the judgment calls, the accountability. And if you try to separate those tasks, you make the entire job role collapse. And junior work is typically those who have the most, let's say, freestanding, independent tasks that can be automated. But the problem here is that the only way to become a senior, advanced, experienced employee is to spend years doing the junior work. You know, the first year legal associates document review is separable, so you can automate it. Same with the junior analyst, you know, financial model or the graduates, I don't know, research deck who they produce for, you know, leaders of the company. These tasks can be lifted out and done by AI really quickly and easily. And then the tasks that resist that kind of separation and automation are those that are senior jobs. Reading, for example, a witness in a disposition and deciding whether to settle on it or pricing, let's say, a deal and managing the board and the stakeholders through that deal. choosing which pitch to take for example into a chief marketing officer and the person that reads it in the room once you're there these are these are very entangled tasks that are heavily dependent and have a massive impact on the relationships or context or consequences and accountabilities that hold the whole job role together so yeah junior work doesn't have that protection and the pattern isn't unique to just software engineers it's happening in journalism in architecture, basically anywhere that entry-level work is primarily about turning information into more structured output versus having to exercise you know really difficult judgment calls with real consequences attached But the problem is is this creates a massive pipeline problem You know, if AI does the junior work, there's no economic reason to ever hire a junior employee. And if nobody hires them, how do they become experienced senior employees of a company? You know, the pipeline doesn't just slow down, it just completely stops. so i guess that's a very doom and gloom view and there are arguments for or against that point of view here and let's say let's start on the optimistic side because it's useful and really important to know so tim o'reilly you might recognize his name o'reilly publications which he got his books at school but he has some really interesting essays now he's been a developer his entire life he has a very very unique and fantastic point of view here and his argument is that he's been watching programming say transition for 40 odd years and a lot of the time these types of panics misread history you know every new programming wave from assembly to to Fortran, to C, to Consumer Basic, to Windows, to Cloud, looked from when you were, I guess, standing in the middle of it, like the end of programming. Every single one of them turned out to be just an expansion of programming. But the expansion wasn't just that the bar got lower, some more of the same people could get involved. Each way brought with it an entirely new skill set and population that previously hadn't been involved. So Fortran didn't make, for example, assembly programmers pointless, it brought scientists and engineers into computing. People would have never got involved with machine code. Consumer basic didn't, let's say, replace C programmers. It brought in teenagers in the bedrooms, a demographic who couldn't have ever been involved in the industry. Operating systems like Windows didn't stop people building applications. It meant that an entire industry of applications was born as a result. So the pyramid got wider at every single level and the people from the previous world didn't disappear. They just moved up. Their skills became foundations that basically others just would build on rather than suddenly this new ceiling that meant that they were screwed and they had to leave the job. Now, personally, I used to buy into this belief and recently I've been wondering whether that is still the case. Now, one argument is the Javon's paradox. So I've talked about this a number of times And this is when something becomes so readily available, whether it's affordable or easy to do, whatever it might be, efficient. Suddenly there's an explosion of that thing. At first, it looks like the efficiency gain from the technology that enabled it would mean that people would lose their jobs. But just as a result of the sheer adoption of a thing, there is so many more applications and jobs as a result. So, you know, steam didn't remove the need for coal. It actually increased it because efficiency made steam power economically viable in applications where previously it would have been impossible to justify the cost. You could apply that to engineering, which is making people 10, 20 times more productive might seem like you need less of them. But over time, the argument is that suddenly there's loads of new tools that you would build that would never have ever been justified or even considered. And as a result, more engineers need it. now where i have an issue with this to be honest is the time scale this argument is basically about where the industry could be in 2035 or 2040 but the pipeline problem i'm really discussing here is about 2028 to 2033 and if there's no juniors in 2025 2026 who are the seniors by 2030 that requires a specific cohort to have been in the job for five six seven years and those people aren being hired You can suddenly have more software getting built in 2035 and needing more employees and hires but having less people to hire because there are just less careers there and the jevon's paradox is essentially an aggregate outcome across the entire economy over a longer period of time so it's the culmination of all these things happening over a 10, 15, 20 year period. But that pipeline problem that I'm actually discussing is about a very specific set period of time. To have seniors in the job for seven, eight years, they need to have been hired seven, eight years ago. Another argument against that, I guess, positive outlook right now, and I know that I've just seen somewhat negative here, is that every previous, say, technology wave, so Fortran, basic, cloud, essentially arrived after the experience of the previous layer had already kind of been established. Fortran programmers had spent years working in Fortran before it then got abstracted away. C programmers brought their understanding of, say, memory management into JavaScript. And each abstraction essentially is being built on the last group of people who had actually done the thing. The wave had come after the experience formed, not before it. But this current wave is just happening so much faster. You know, something really is happening to entry-level work faster than any previous transition in history. Although the Jevons paradox in the long run, 15, 20 years period of time, might be true. There might be just such an explosion of things over this next long period that eventually there's more jobs. But over this shorter window, companies are making a massive mistake by not thinking about their junior pipelines. So the Jevons paradox argument that's very optimistic and the pipeline problem that I'm really discussing here aren't competing. It's kind of like a sequence of events. There needs to be an expansion of demand, but the expertise just isn't there anymore to meet it. And you eventually get that shortage that I mentioned about Boeing facing in the aerospace industry. And that'll happen in 2035 rather than in 50 years later. But it will happen. See, I guess to conclude here, the argument should say that each transition of new technology and, say, programmers and software development really did expand the surface area, the amount of potential in the space. And that is a historical analogy that will probably play out to be true. But the problem right now, and this is why I think companies are making a massive mistake, is that the cohort that will become the leaders of a company in 2032, 2032 and 2035 aren't being hired and therefore won't be trained. And there are companies that are doing the right thing here. IBM tripled its entry-level intake in 2026, one of the very few companies that have done this. Publicist, the marketing group, tried to restructure its graduate training program around AI fluency rather than just getting rid of it. But these are a drop in the ocean in terms of the general statistics across the market right now. And that billboard that people put up in Seattle, it's a joke, but it's a reality now, 50 years later. The aerospace companies who made those cuts are now probably looking back, really regretting some stupid decisions. And many of these big, big companies, especially, are going to do the same and are going to make the same mistake in the future. Anyway, thank you for listening and I'll see you next week.
