Meh... just rehashing what he said before. The paper itself is fundamentally flawed, examining only a minuscule portion of the job market. If we step back and look at Europe's struggling economies over recent decades, we see that economic downturns disproportionately affect young people. Greece serves as the poster child, followed by Spain and Italy. In Germany alone, we've lost 50,000 jobs in manual labor heavy industries (mainly automotive) this past year. We're also seeing a 60% decline in apprenticeships for labor intensive roles at DAX companies that aren't even AI affected yet. AI has become a convenient scapegoat for a faltering economy driven by geopolitical tensions, protectionism and unqualified leadership in the world's largest economies. Roaring 20s indeed.

For a startup, I now can solo it with the help of AI far beyond what was possible a few years ago.

This does not mean I otherwise would have hired a few juniors. More likely I would not have taken that business beyond the idea stage.

But direct massacres I can personally observe are in translation, copywriting and illustration. Yes, a good person can do better, but for a majority of cases AI has already become "good enough".

Then there is an onslought from just non-AI automation, in e.g. retail and finsncial services. This had been building for over a decade, but the pandemic lockdows exponentially accelerated it.

Don't underestimate the effect of Elon nuking the Twitter employee base when he acquired the company. All the predictions were that Twitter would die. It hasn't. That sent a powerful message to the other leaders of tech companies: "You can do just as much with far fewer people" / "Most of your employees aren't doing needle moving work". Shortly thereafter, Zuck announced layoffs at Meta. These new attitudes towards headcount have cascaded down from the top-tier tech companies down through the ecosystem of companies.

Young people from where? I’m in South American and the young people here are earning much more than their parents earn but 1/3 what software engineers earn in San Francisco writing software.

Interest rates have way more impact.

By what mechanism can you distinguish a decline from one cause compared to another?

I could understand if, given enough time for all of the factors involved to occilate that you could pick out a signal, but that's not the data that exists right now. Surely the only way to identify the first time instance of a cause of decline could only by accumulating a count of clear instances where the cause occurred and measuring those as a proportion of other cases.

In Canada, mass immigration is creating high unemployment for young people. Don't be so quick to blame AI when government policies are to blame. https://www.kingstrust.ca/wp-content/uploads/2024/11/TKTC-De...

Programming in a structured high-level language like C is much more productive than programming in assembly. Programming in a managed language with GC can be more productive still.

The creation of package managers and the widespread availability of open-source repos means developers don't need to write as much from scratch.

The creation of search engines and Stack Overflow did (and still do) much of the useful things that people use AI for (boilerplate, debugging obscure error messages).

Machines have gotten exponentially faster for the last several decades. This means devs need to spend less time optimizing code. And the time to compile and run speeds up, meaning you can prototype things faster.

Why is it, that somehow none of these inarguable improvements to the speed and efficiency of development haven't lead to a a massive decrease in the number of developers? If we take it as axiomatic that AI significantly improves productivity, why, for the first time in history, does that not result in more programming jobs?

Researchers need to project the pre-Covid hiring rates and omit the subsequent overhiring boom to credibly report on any believable loss in jobs. Dropping personnel who should never have been hired to begin with is a return to normal, not a new concerning trend. Start by showing the data from at least 2015 onward rather than lying with statistics using charts starting in 2021 for maximum drama.

> For automative occupations, a lot of it is software engineering, auditing, and accounting, where there are well-defined workflows and LLMs are good at doing one-off tasks without a lot of feedback. For augmentative cases, you’re looking at more complex or managerial roles.

Nope coding is definitely augmentive for Sep 2025 and before. Maybe more so than managing.