What's going to happen to junior engineers?

In the 2010s, companies were hiring a lot of juniors. This was partially an economic decision: companies were hiring everyone, and juniors were a lot cheaper and easier to hire than senior engineers. In 2025, this has mostly dried up. Tech company hiring is much less common and has tilted towards seniors, who are seen as less risky hires. The economic situation will eventually change. But I’m worried about AI.

Why is AI a problem for junior engineers?

Frontier AI models are pretty good at software engineering. Fortunately for me, they’re not as good as senior engineers yet: they struggle with very large contexts and deep familiarity with specific codebases that aren’t in their training data. If you’re a strong engineer, you can make better high-level decisions than AI models can, and so add value. I can throw a task at an AI agent and tell at a glance “yep, this is what I would have done”, or “no, this is over-complicated, try again like this”.

I work with AI models like the chess players of the late ’90s worked with chess engines. Together, we’re more capable than either of us would be apart. But I don’t know if I could have done this as a junior engineer. At that stage in my career, I was more playing the role of the AI model: working away on problems and occasionally being set straight by my more senior peers. What niche can a junior engineer fill, now that AI can play that role for pennies on the dollar?

Without juniors, where will the seniors come from?

One argument is that it’s in companies’ best interests to hire and train juniors anyway, because there’s no other way to get seniors. Tech companies never expected much productivity out of fresh juniors anyway - the point was always to invest in their growth. But tech companies used to start getting net-positive value out of juniors relatively early (i.e. when they became able to work on well-scoped tasks with frequent supervision). If that niche is being filled with AI, tech companies will only start getting value when juniors get enough expertise to supervise the AI models themselves - after another six months to a year. In the current tough economic environment, that might be a hard sell.

In general, the argument that companies have a long-term interest in hiring junior engineers does not have a lot of predictive power over what they will actually do. Very few of the decision-makers at tech companies expect to be in the same role a decade from now. I don’t expect them to take a hit to their own careers in order to improve the general state of the industry.

Will juniors be better at AI?

Another argument is that AI-native juniors who grew up using ChatGPT will be better at it than current seniors, and will be able to offset their lack of productivity with more AI-specific skills. I don’t really buy this. When people talk about “tech-native” generations, they’re talking about the mean: e.g. how good is the average millenial at using a PC compared to the average baby boomer. Engineering professionals who pay attention to new technologies will become expert in them, as they always do.

I also don’t think that the skills at using AI to do software engineering are all AI-specific. Don’t get me wrong: there are some AI-specific skills involved. But being good at working with AI requires you to be good at designing software and reviewing code more than it requires you to be great at prompting.

Companies might hire juniors instead of seniors

Is it possible that the dynamic might work the other way around? Could companies decide that they don’t need expensive senior engineers when AI does most of the work, and instead hire juniors to pair with the AI models? I don’t think so.

As I said earlier, the key skill to working with AI is knowing better than it about a couple of key points - for instance, knowing when the solution it’s working up just looks way more complicated than it needs to be. A less experienced engineer just isn’t going to be able to tell when that’s happening.

Some companies might try this. But I don’t know what would prevent those companies from skipping the engineers entirely and just vibe-coding everything themselves. If you’re the type of executive who thinks you don’t need human engineering expertise, aren’t you also the type of executive who’s reluctant to rely on junior engineers?

AI might increase the amount of engineering work

The most optimistic argument is that AI-driven development will increase the amount of software in the world dramatically, and that will lead to enough demand for engineers that much of it will spill over into the junior market. For this to be true, AI would have to land in a “sweet spot”: powerful enough to generate a ton of business value, but not quite powerful enough to be able to maintain its own software independently.

Of course, in this scenario most of the demand will be for seniors that can effectively manage the AI. But if the market is tight enough, companies will try and hire skilled juniors to pick up the slack, or to take the time to train juniors up until they can do the work. I would like this to be the case. But I do think it’s more likely that either (a) AI fizzles out and remains a useful developer tool, or (b) AI is powerful enough to do the majority of engineering tasks, and most engineers are out of a job.

Is AI really that useful?

I want to pay lip service to a final argument I’ve heard: that AI really isn’t that useful and won’t have any real effects on software engineering, so juniors will be alright once the market conditions turn and companies start hiring again.

I don’t know how you can be a working software engineer at a normal tech company and believe this¹. Right now AI is so clearly useful to my everyday work that even if progress stalled today it’d still be a transformative technology (on par with the debugger or Docker, say). But I do want to acknowledge that I could somehow be wrong.

Final thoughts

To put it mildly, a cool market for junior engineers is not good for the industry. Even aside from long-term planning, I personally really like working with juniors. It’s just nice to work with smart, motivated people who are new to the industry. I really hope I’m wrong about all this.