Everyone is cheering that AI turns you into a "full-stack everything." But no one is questioning what it means when all of us are full-stack everythings.

Here's the uncomfortable math: if AI hands every developer the same superpowers, those powers aren't superpowers anymore. They will become basic requirements.

The Essay That Hit a Nerve

A post titled "LLMs are eroding my software engineering career and I don't know what to do" went viral in developer communities, garnering hundreds of passionate responses. Its core message was direct: AI is diminishing the value of specialization, turning everyone into a generalist.

It wasn't the essay that surprised me. It was the reaction to it. Senior developers, with 10-15 years of experience, publicly sharing their career anxiety that their skills might be commoditized instantly.

The "Everyone Is 10x" Lie

The common story is like this: AI can make each developer ten times more productive. You can now work on backend, frontend, infra, data — everything. You are a generalist superhero. 🦸

But economics doesn't care about your productivity fantasy. The reality is:

→ AI gives every developer the ability to do generalist work
→ The supply of "generalists" explodes
→ Oversupplied skills get paid less, not more
→ Companies need fewer people to ship the same output

This is not a theory. It's about how much is available and how many people need it. The exact same principle that renders AI-generated stock images completely valueless is beginning to affect commoditized development work.

When everyone can build a CRUD app with a prompt, building CRUD apps pays CRUD wages.

Specialization Was Your Moat

Consider the reasons why senior engineers were paid more. It wasn't because they were capable of "doing a little of everything.” It was because they possessed depth — whether in distributed systems, performance engineering, security, or domain knowledge that required years to acquire.

That depth was a moat. It was hard to replicate, hard to hire for, and therefore valuable.

The pitch now is: "You don't need to worry about depth. AI will fill in the gaps." However, if everyone's gaps are equally filled by AI, your moat has been destroyed. You have gone from being "the person who truly understands X" to "one more dev with a ChatGPT subscription."

That's not a promotion. That's a demotion with better tooling.

What Actually Holds Value

I believe the solution is not to refuse AI. That's just denial with extra steps. However, I am convinced that considering "AI turns everyone into a generalist" is a harmful idea that only advantages employers over engineers.

What still holds value in a world of commoditized code generation:

→ Taste — knowing what to build, not just how to build it
→ Domain expertise — understanding the problem space deeply enough to ask questions AI can't
→ Systems thinking — seeing second-order effects, failure modes, and tradeoffs that don't show up in a prompt response
→ Judgment under ambiguity — making the call when there's no clear right answer

None of these are things you develop by becoming a "generalist who vibes with LLMs." They come from going deep on hard problems over time. 🔍

The Real Risk Nobody Talks About

The scariest part isn't that AI replaces developers. It's that AI makes developers interchangeable.

If what you are offering as your value proposition is “I can ship features fast with AI assistance.” Well, congratulations! So can the next hundred applicants. You’ve essentially optimized yourself right into a commodity.

Companies love commodities. Commodities are cheap and replaceable. That's great for hiring managers. It's terrible for your salary negotiation.

The developers who will thrive aren't the ones who become the best users of AI. They're the ones who remain the hardest to replace even with AI. That means going deeper, not wider. It means building expertise that can't be prompted into existence. 💡

The generalist hype is a sugar rush. Specialization is still the long game.

So here's my question: are you building skills that survive commoditization, or are you just getting faster at work that's about to get cheaper?