67
submitted 1 week ago* (last edited 1 week ago) by Septimaeus@infosec.pub to c/experienced_devs@programming.dev

ETA: summary…

Author describes how agentic coding tools have disincentivized the developer training and mentorship pipeline that the software industry has implicitly relied on to produce seasoned developers, predicting that once the shortage is fully realized it could take years to reverse.

The solution proposed is for managers and senior developers to get ahead of this shortage by foregoing the immediate yet shortlived productivity maximization offered by the agent-multiplied-senior structure in favor of longterm sustainability by deliberately facilitating experience fan-out to juniors.

The methodology proposed is based on a successful nuclear engineering apprenticeship program carried out on active service naval submarines.

you are viewing a single comment's thread
view the rest of the comments
[-] minfapper@piefed.social 2 points 1 week ago* (last edited 1 week ago)

It makes me wonder how these 5 senior engineers came to found a company that works like that. Do they not like writing the code? Do they have dreams of being rich? Do they believe the product they will produce is so essential that the process doesn’t matter?

I started programming when I was 14 years old, and turned 34 this year. That makes about 20 years.

Figuring out the syntax of a new language, the satisfaction of getting something to compile, and getting the "perfect" dev environment on my Arch Linux install was a lot of fun. I wouldn't trade that experience for anything, but at this point I've gotten all the fulfillment I'm going to get from it.

Even before AI agents came along, I had already moved on to caring about the impact of the software: benefit to society (for my open source contributions); revenue to company (for half my day job); learning business skills to improve my earning potential (for the other half of my day job).

I would delegate the actual coding to junior engineers, and spend most of my time designing the higher level architecture and everything else needed to herd them into delivering the high level business objective.

Unfortunately, the current generation of AI agents (Opus 5, GPT 5.6 Sol) can do a somewhat similar role to those juniors. They don't have the human element that made it fun to mentor the juniors, but they also don't have feelings so I can give them feedback without worrying about coming off as a condescending asshole.

Dunno if that's helpful at all to you, but if you're genuinely wondering why a decent chunk of senior engineers are sticking around in the AI world, then there's one perspective.

The only way I see this ending is Juinor wages plummeting and turning into unpaid internships or paid (the other way) ‘educational programs’.

I do have a slightly more positive note on that. Universities are slow to adapt their curriculums to new technologies, but they will eventually care that they're producing graduates that provide so little value to companies that they're unemployable.

They'll begin to produce graduates that are the equivalent of today's mid level and senior engineers. While I don't know exactly what kind of education can instill in 4 years the essence of what took me decades to learn, I do know that pretty much every generation of humans have surpassed their parents so far. So I have every confidence you kids will figure it out.

this post was submitted on 25 Sep 2026
67 points (93.5% liked)

Experienced Devs

5730 readers
10 users here now

A community for discussion amongst professional software developers.

Posts should be relevant to those well into their careers.

For those looking to break into the industry, are hustling for their first job, or have just started their career and are looking for advice, check out:

founded 3 years ago
MODERATORS