Programming

I tried reading the article, got about halfway through and it just keeps repeating the same thing over and over again. It starts with this:

There’s a shortage coming that nobody’s put on a slide yet: senior engineers. Every company on earth is about to want one who can drive a fleet of agents with judgment — and almost nobody is making new ones anymore, because AI ate the entry-level work juniors used to learn on

That's it. That's the whole article as far as I was willing to read. There's almost no extra information further down. Holy hell.

Unemployed principal software engineer here. This graph is a lie.

Didn't read the article but I love the braindead chart:

LLMs are a product that destroys its supply chain — unnamed Microsoft exec

So soon? I figured it'd take a couple years for them to realize how they shot themselves in the foot by not hiring juniors anymore.

There’s a terrific writeup from the folks at FlintAI — their team is now 5 devs and 45 agents, all living in Slack. Nine agents per developer. The agents take roles — triage, deploy, review — the human steps in for the two decisions that actually matter (code review, release approval), and they report something like a 3x productivity bump. It’s a genuinely impressive piece of engineering and you should read it.

That’s the model everybody’s racing toward right now: take your best seniors, strap nine agents to them, and harvest the multiple. And it works! I’m not knocking it (and I’m doing it too)!

I'm a senior but if the future of my work is reviewing AI code all day I will buy a bunch of goats and become a goatherder.

You Can’t Just Buy Them

I can hear the objection, because it’s the obvious one and I’ve made it myself: just hire seniors when you need them. Let the other guy train them. Poach the finished product. It’s cheaper, it’s faster, and it has worked for the entire history of this industry. Right? WRONG.

Here’s why that breaks: everyone is about to have that exact same idea at the exact same time. Every company that quietly decides juniors are too expensive is also quietly deciding to buy its seniors on the open market later.

One might think that companies really ought to be smarter. But the histories of whaling and fishing shows that profit-oriented industries are absolutely capable of completely and irreversibly wrecking their own economical foundations for a few years with a bit of extra profit.

Stop hiring and mentoring junior engineers and you’ve quietly unplugged the machine that makes senior ones. A few years out, you run dry — right when everyone else is dry too, and the price of the few left goes through the roof.

Written by AI

IMHO. This haas been common sense for most senior devs for a while now. Looking at AI workflows there is little opportunity to build up the skill and experience required to become a senior.

This was written by ai, right?

Companies are not only no longer taking on juniors, they're making hardware prohibitively expensive. Where do they expect to get their talent if they force everyone to use their braindead toys?

This post is the piece those two were circling around: what happens to the bottom of the chart, and whose job it is to fix it. Spoiler — it’s yours, senior engineer. It’s mine. And most of us are about to get it wrong in a very specific, very tempting way.

BS. Most good seniors I know would love to hire more juniors, feel badly for the mass of recent college grads who are unemployed, and enjoy mentoring new engineers without needing a wall of AI slop to tell them to do so. When you give us control over the hiring budget and let us hire juniors again, you can blame us for the future senior engineer shortage. Until then, blame the people actually at fault (management, executives, the investor class), not us.

Everyone thinks LLMs boost productivity. But from psychological point of view, the developer being both the subject and the object is no where near scientific and reliable.

In other words, the developer who is using the LLM also reports they are productive. But humans can rarely be good judge over themselves. That's why we've got therapists. So we have to wait for some serious and well done research if LLMs boost productivity.

Another concern I have which prevents me from using LLMs, is the unknown long term effect on the codebase. You know, writing something is much easier than maintaining it for years after years. And every decision you make, you must be mindful of this. So I'm gonna wait and see what'll happen after 1-3 years to the softwares with heavy LLM assists. Will their codebase still be easy to maintain?