If you or your friends are currently hunting for your very first job, you’ve probably noticed by now that it’s not that easy.
The German economy isn’t doing too well, so companies are happily sitting on their hands: no new hires, or only bringing in seasoned pros. And if someone on the payroll is „too seasoned“? Well off they go, take that early retirement package. Layoffs? Bad for the company image. And as for image campaigns on Instagram? Forget it. No budget, and besides, we don’t have underpaid Gen-Zs around anymore to run them.
But what if there’s actually more to it? A couple weeks back I was listening to a tech podcast. Mike Krieger, the instagram co-founder who now works at Anthropic (one of the major AI companies) was asked about a comment Dario Amodei – CEO of Anthropic – made about entry-level jobs. He claimed that up to 50 percent of entry-level white-collar jobs (dt. z.B. Bürojobs) could vanish in the next one to five years, possibly spiking unemployment to 10-20%.
Now, at first glance this sounds like a typical Silicon-Valley fantasy (doomsday by power-point or something). But asked about this, Krieger admitted: Even inside Anthropic, tasks that used to be done by junior engineers are already being handed to Claude, their AI system. And that they are already hesitant to hire juniors at all.
As someone who studies Business Information Systems and might end up in a software development firm, this sparked my interest. I started looking for data to back this up.
You are the canary
A new Stanford study tracked millions of U.S. payroll records to see what is actually happening to entry-level jobs.It does not look great for 22 to 25 year olds. Since late 2022, employment for young workers in the most AI-exposed jobs, like software developers or customer support has fallen by about 6%, while older colleagues in the same jobs have seen employment by 6-9%. Overall, there has been a 13 percent decline in employment in occupations that are highly exposed to AI.
But here’s the thing: the study highlights the difference between automation and augmentation of jobs. Job automation means that tasks are replaced with AI, augmentation means that AI enhances your work by making you more efficient. And the study shows that when AI is used to automate, entry-level jobs shrink. But when it’s used to augment, these jobs actually grow. In other words, whether you’re in trouble or luck depends on whether the AI is replacing or helping you.
But why are entry-level jobs the first to be affected? Often, junior employees start their careers handling routine tasks: updating spreadsheets, preparing reports or drafting presentations. These are precisely the kind of activities that GenAI systems like ChatGPT now can perform quickly and reliably. Older colleagues in contrast are simply more experienced and have practical know-how and contextual understanding, which is hard to codify and replace.
In other words, AI is raising the bar for junior employees. The “easy” tasks that once served as stepping stones are disappearing, leaving us — the next wave of graduates — needing to prove ourselves in more complex ways right from the start.
The researchers call these young workers the *canaries in the coal mine*. Just like the birds miners once carried underground to warn them of toxic gases, early-career employees are the first to feel the effects of AI. If their jobs are disappearing now, it could be a warning sign of what’s coming for everyone else. Great! So we’re the canaries, getting thrown under the bus before our careers even start. As if the climate catastrophe, the war in Europe and looming recession weren’t enough, we might not even get a job to begin with.
Okay, that does sound pretty grim. Sadly, the situation in Germany might not be better: A recent YouGov-survey found that about one in three Germans already fears losing their job to AI.
Which jobs are affected most?
So where do we see entry-level opportunities disappearing the fastest? The biggest cuts to entry level jobs show up in IT, consulting, finance, law, marketing and administration. Training spots in software development and tax consultancy have dropped by about a quarter, consulting firms are hiring fewer graduates, and law offices rely more on AI for research. Even routine marketing, customer service, and admin tasks are increasingly automated
The Frankfurter Rundschau points to a paradox: while entry-level positions are vanishing, the demand for highly skilled and experienced workers remains as strong as ever. Great news if you’re already senior, not so much if you’re just trying to land your first job.
What now? AI Literacy?
Krieger (the guy from the podcast) also thinks that there is „a continued role for people that have embraced these tools to make themselves, in many ways, as productive as a senior engineer“. But there is more to it than simply using these systems. Everyone can type a query into ChatGPT and call it a day. I think it’s more about knowing when to use what AI system, when to trust the AI system, and when not to use AI because it will just make something up, or perform worse than you.
That’s basically what AI literacy is: understanding what AI tools can and cannot do. And just like being proficient in excel was once the universal baseline skill for office work, AI literacy might turn into the new must-have-on-your-resume-skill.
The good news is: You can actually learn this. The UDE already offers formats like Zahlen, Daten, Fritten at the DataCampus, where experts talk about data and digital skills, and there are workshops and online courses on prompting and AI basics. (find more information here)
