Hiring has not yet seen the broad collapse that was forecast.
When looking at hiring changes related to AI, it’s easy to get caught up in quotes from tech leaders like Anthropic CEO Dario Amodei who predicted that artificial intelligence could wipe out roughly 50% of all entry-level white-collar jobs within one to five years. If this occurred this would push unemployment possibly above 10%, this being strictly in recession territory.
When I spent some time reading about the impact of AI on the job market however, the picture becomes quite a bit more muddy. This post is my attempt at trying to untangle different claims, research and frameworks about the impact of AI on job markets.
What’s reported now
There appear to me to be three things to balance when looking at aggregate changes to the job market:
Changing interest rates: the market for software related hiring driven by low interest rates pre-2020 was remarkable. After interest rate increases, companies began layoffs which has caused workforce churn which is extending until today. It’s difficult to draw causality between whether layoffs and restructuring are plans built pre-ChatGPT in 2022 which are still being enacted in 2024/2025 or whether it’s truly a phase change because of AI. The Yale Budget Lab which monitors AI impact to labor markets has seen no aggregate change1.
The data-centre boom: There is undoubtedly a boom in construction of chips, servers and all sorts of demand for HVAC workers and electricians. This is causing a net increase in jobs from AI even if you suppose there is some job loss to white collar workers2.
Doom and gloom for young workers: Even though the data does not appear in aggregate, young workers are not entering the job market with good vibes3. It seems reasonable to assume that a drop off in hiring for early career computer science majors would be caused by AI, but it’s unclear whether the first affect I mentioned (higher interest rates) means this is simply a continuing trend. The data here is mixed and there are some arguments that we will need to wait some time to see4.
Forecasting
The above three effects are contradictory and significantly muddy the picture for anyone attempting to perform an analysis of job markets today. It appears clear that we can say an “apocalyptic” scenario for jobs will not appear in the next 1-2 years with the forecasted rate of model growth and the hiring picture we see today. It’s possible even that demand for “AI Engineers” begins replacing the traditional coding role we saw.
Personally, I don’t take anything away from the current job numbers. I think if you are genuinely interested in computer science, the reality is the career will be incredibly different to what it was in the past. At the same time, this could have been said for any cohort of graduates. The unique conditions of hiring between ~2015-2020, where employment was near guaranteed are also gone. This has nothing to do with AI however and mixing these up makes the picture look unnecessarily gloomy.
We will look at an economic model in more detail to think about forecasting in the next post.
https://budgetlab.yale.edu/research/tracking-impact-ai-labor-market ↩
https://www.economist.com/finance-and-economics/2026/09/04/the-jobs-apocalypse-is-postponed-an-ai-jobs-boom-is-here ↩
https://economist.com/finance-and-economics/2026/05/13/is-ai-putting-graduates-out-of-work-already ↩
https://www.brookings.edu/articles/research-on-ai-and-the-labor-market-is-still-in-the-first-inning/ ↩