AI is taking over junior IT positions

Artificial intelligence is changing the job market. It’s also changing the face of technology professionals. It’s also making things more difficult for IT professionals, at least when it comes to finding some types of junior-level jobs.

This is according to a study from Stanford University, which examines the overall impact AI is having on the labor market using data from the United States over the past three years as a sample. Its conclusions indicate that the employment rate in positions most affected by AI has fallen by 13%. Among the most affected positions are administrative assistants and accountants, but also developers and entry-level jobs in the IT sector.

This isn’t the only study that points in this direction. A calculation by The Washington Post also concludes that a quarter of programming jobs have disappeared in the past two years. It draws on official US employment data. These are, however, entry-level positions, more basic jobs that serve as a gateway to the labor market. As the Post points out, programmers are — followed by developers, IT technicians, and software quality testers — among the professionals most threatened by generative AI.

While finding your way into these junior positions is difficult, overall labor market trends have remained relatively the same, and things have gone well for more experienced workers, according to Stanford’s research data. This leads the study’s authors to conclude that, rather than helping the workforce do things better or more efficiently, AI is taking over those jobs that can be automated with artificial intelligence.

Some of the market movements seem to confirm the perceptions of the Stanford researchers. Major tech companies have launched a wave of staff cuts, in which AI appears to play a significant role. Microsoft CEO Satya Nadella has even indicated that 30% of Microsoft’s code is now written by AI.

Use for work support

Paradoxically, what usage studies tell us is that, in most cases, what teams are doing is using AI to improve how they work rather than having it do everything for them. This is confirmed by 57% of professionals. “Usage leans more toward augmentation — such as having AI review your work, asking it to teach you things, or repeating work — rather than automation,” Alex Tamkin, a researcher at Anthropic, told the Post.

This is despite other analyses indicating that, at least in software development, AI slows down human work because it actually makes completing actions take longer, despite the perception that it should be speeding up developers’ work.

“This gap between perception and reality is striking: Developers expect AI to speed them up by 24%, and even after experiencing the slowdown, they believe it has sped them up by 20%,” METR, which evaluates AI models, tells NBC.