Things I Currently Believe About AI and Tech Employment

Hey, I could be wrong!

People love to ask me what I think is going to change about engineering management and technology careers given the current AI innovation boom. I am flattered to be asked, but I am not sure I have a good answer. Things keep changing so much in the space, both from real innovations (like Claude Code) and promises of near-miracles (like 90% of all code being written by AI sometime this year), it’s hard to keep up or know what’s true. So, while I have some direct experience, and I continue to collect a lot of interesting anecdotes, I think we’re still in the realm of “futurism” when trying to guess what will happen.

So, with that out of the way, here is one of my current best guesses. This is not based much on the current state of innovation, except to take as a given that AI is proving to be measurably supportive of building software. It's instead based on my understanding of economics, human nature, and the skills of people working in tech.

So long as there are software engineers, there will still be a range of junior — mid — senior engineers.

Some people say that AI will get rid of the need for “junior” engineers, generally taken to mean “people with some training but no full-time software engineering work experience.” Others say that AI will get rid of the middle-tier engineers, those with a few years of experience but who aren’t yet senior. And there’s probably some who think they make even very talented senior engineers obsolete, but more who just really want them to make very talented senior engineers cheaper to hire. But will either of these come to pass?

Mostly Junior?

It’s clear that AI has been a massive accelerator of what has been called “low-code” or even “no-code.” I know of startups that are hiring nothing but junior people to act as prompt coders, with the senior experienced tech founders overseeing. But these are very early-stage companies, with greenfield codebases. Clearly AI is making it much easier and cheaper for them to get started, and you can now build some types of software early-stage businesses without mid-career engineers. But are you going to keep these “junior” engineers junior forever? Churn them out and hire new ones the minute they develop the context understanding of your business that makes them a) more productive and b) want more money/promotion?

If you believe “junior engineer” is now in a category other than “knowledge work”, one that has a workforce that can be staffed with a lot people who need minimal training to do the job and far fewer above them to oversee the work, maybe this is true. But it doesn’t sound like we’re at the cusp of turning this job into a largely lower-skill position yet. I’d sooner believe that prompt/vibe coding becomes a skill that other knowledge workers need to have and we just have much smaller engineering teams overall than engineering shifts to primarily juniorization.

Mostly Senior?

Others say this means you don’t need junior engineers. By most accounts I’ve heard, AI coding tools are making a huge difference in productivity of senior engineers, which naturally slows down the demand for hiring at all. And it’s tempting to delay hiring of junior engineers, the ones that you have to invest in before they become productive, in order to take full advantage of that AI boost for the existing employees. But over time, if the jobs continue to exist, we need new entrants to our talent pool to replace leavers. Not to mention all of the other positive effects bringing new talent into a team creates. I’ve written about this already, and I stand my ground: if you think you will continue to need to hire engineers into the future, you should really keep hiring junior ones.

What about engineers in-between “Junior” and “Senior”?

So long as there are junior engineers, and senior engineers, there will be career paths with midpoint levels. The description of skills might change, the compensation potential might change, the total employment might change, but I am skeptical that one role will disappear but the other two will remain. The only alternate future I can imagine is one where you may start people as a junior engineer but then everyone specializes above that; you hire them as “analysts” and they figure out a business specialization that provides the career path rather than the technical career path. My skepticism about this “full stack business person” is that, well, most people just aren’t good at everything. We tried to get rid of just developer specialization with “full stack” roles, but at scale, this has yet to be widely successful. It turns out, specialization is valuable, and can be more valuable than being a middling generalist.

Corollary: Will Efficiencies Kill Demand?

So now we have to talk about what this will do to the overall demand for engineers. Will it plummet? Is the job category of “people who build software-based products” going to have the same massive decline as, say, manufacturing? It helps to know why manufacturing employment declined in the way it did. While there were many factors, we know that manufacturing got both much more efficient, and the demand for manufactured goods slowed.

Taking it as a given that AI will meaningfully improve the productivity of producing software-based products, do we think we will also saturate the demand for those products? Or will the demand continue to grow? Time will tell. The biggest AI proponents predict that efficiencies will reach some point that the AI satisfies them completely, so I guess demand doesn’t matter in that case. But I’m skeptical.

Engineering is changing, but probably not going away

I am not a futurist. I don’t believe yet that we are going to achieve “AGI” (whatever that means). I do believe you can, and people do, “write” a lot more code now via prompting, which is a step beyond the elaborate auto-complete copilots of even a year or so ago. How will this work over time, with many people, in increasingly complex systems? No one knows for sure, but my current bet is that specialization still matters, it still produces marginal utility, and thus the specialization of software engineering, while changing, will stick around.

No AI was used to write this post (unless you count spell check!)

Enjoy this post? You might like my books: The Manager’s Path, and Platform Engineering: A Guide for Technical, Product, and People Leaders, available on Amazon and Safari Online,