AI for Automation
Back to AI News
2026-03-22AIdeveloperscraftsmanshipwork cultureautomation

AI didn't kill craftsmanship — the market did

A trending HN essay argues AI tools aren't the problem — employers force developers to use AI or fall behind. The market, not AI, kills the craft.


A software engineer in Seoul just published an essay that's struck a nerve across the developer world. Hong Minhee's "Why craft-lovers are losing their craft" argues that AI coding tools aren't the enemy — but the market pressure to use them is destroying something irreplaceable. With 57 points and 39 comments on Hacker News, it's the kind of quiet, uncomfortable truth that keeps showing up.

The core insight: nobody actually banned hand-coding. You can still write every line yourself. But if your coworker ships 3x faster with AI, your manager doesn't care about your beautiful, hand-crafted solution. The market made loving your craft a career risk.

Hong Minhee, Seoul-based software engineer and open-source maintainer who wrote the viral essay on AI and craft alienation

Hong Minhee, the Seoul-based developer behind the essay

The invisible divide AI made visible

Minhee draws on two contrasting developer perspectives that AI coding assistants suddenly revealed:

Developer A: "The puzzle didn't disappear — it moved to a higher level." AI is just another step up the abstraction ladder. No loss felt.

Developer B: Mourns losing "the feeling of holding code in our hands and molding it like clay." The process itself was the point. AI skipped the part they loved most.

Before AI assistants, both developers looked identical — same desks, same keyboards, same output. The tools made an invisible motivation gap suddenly impossible to ignore.

Why the market — not AI — is the real threat

Here's where the essay gets sharp. Minhee applies a concept from economics called "alienation from the act of working" — the idea that when your job becomes purely mechanical, you lose the distinctly human part of creation: imagining something and deliberately shaping it into reality.

The mechanism works like this:

Nobody forbids you from writing every line by hand. But companies measure you against AI-assisted colleagues. Developers report adopting these tools not from desire, but from employment necessity. The tool is neutral — the market weaponizes it.

As Minhee puts it: developers who invested their identity in the process of creation feel the loss most deeply. Those who always cared about results barely notice the change.

The escape route that actually exists

Minhee offers a real counterexample from personal experience. As a full-time open-source maintainer funded by grants — not traditional employment — there's no productivity comparison pressure. In this setup, AI handles boring scaffolding work while meaningful creative work stays human.

Same tools, completely different experience. The difference isn't the AI — it's the economic structure surrounding it.

Who should pay attention

  • Developers feeling "AI guilt" — if you feel pressured to use AI tools you don't enjoy, the essay validates that your frustration isn't irrational
  • Team leads and managers — measuring developers purely by speed may be destroying the craftsmanship that produces your best, most maintainable code
  • Anyone in a creative field — this pattern isn't unique to programming. Designers, writers, and artists face the same market pressure to automate the parts of work they love most

The uncomfortable question

The Hacker News discussion with 39 comments echoes a broader pattern: multiple commenters argue that interest rates and over-hiring — not AI — drove most recent tech layoffs. AI became a convenient excuse for decisions companies would have made anyway.

One senior engineer shared concrete examples of AI-generated code containing bugs that doubled compute costs — redundant async tasks that nobody caught because the code "looked right." Speed without understanding isn't free.

The essay's conclusion is uncomfortable but honest: don't blame the tool. Blame whatever forces people to use tools they don't want to use, on terms they didn't choose. The question isn't whether AI can write code — it's who gets to decide how the productivity gains are distributed.

Read the full essay by Hong Minhee.

Related ContentGet Started with Easy Claude Code | Free Learning Guides | More AI News

Stay updated on AI news

Simple explanations of the latest AI developments