HN suspects an ACM essay on AI and coding was itself written by AI
The essay's real argument: the hard part of programming didn't shrink, it relocated — from recall ('how do I write this') to judgment ('does this actually make sense'). One programmer's response summed up the thread's mood: "How many times can we have the same discussion."
One HN veteran flatly rejected the essay's framing: coding did get easier, he said, he just never keeps the gains, because competitive markets absorb them. A freelance product engineer made the opposite complaint from the same thread: his codebases have never been healthier, and he's never enjoyed the job less.
The bluntest read borrowed an old parable: a farmhand who'll chop all the wood you want, then quits the moment he's asked to sort a basement of potatoes.
Why CACM says the hard part moved from typing to judging
The essay, published in Communications of the ACM's Opinion section, makes one narrow claim: coding didn't get easier or harder in some blanket sense, it changed which muscle gets tired. Its central line, as quoted in the HN thread that followed it: "the hard part moves from recall ('How do I write this?') to judgment ('Does this actually make sense?')." Syntax memorization, boilerplate, API lookup, the grind junior developers used to spend years on, is now largely automated. What's left, per the essay, is the harder-to-teach skill of knowing whether the automated output is correct, maintainable, and worth keeping. Its subtitle states the stakes directly: "The future of software development will belong to those who can think clearly at scale, maintain durable mental models amid rapid change, and integrate machine-generated output into human-directed intent."
That reframing sits between the two loudest camps in tech commentary right now: AI ended programming, or AI is a toy for boilerplate. The essay's claim is narrower, and on Hacker News it turned out to be the more provocative one: nothing got easier, the job just changed shape.
The parable HN reached for: a farmhand who quit over sorting potatoes
One reply borrowed an old Alan Watts story to make the essay's point concrete: a farmhand saws more logs and mends more fences in a day than anyone the farmer has ever employed, then quits the moment he's asked to sort a basement of potatoes into what to sell, what to seed, and what to throw away.
"It's all right mending fences and chopping wood, but this potato business is decision after decision after decision." — the commenter who reached for the Alan Watts parable
Physical labor has an end state. Judgment doesn't, and that's the tax several replies said AI actually moved onto programmers: it took the log-sawing and left them in the basement with the potatoes.
Is the essay about AI written by AI?
Not everyone accepted the essay on its own terms. One reply didn't buy its polish at all.
"I always find it hilarious when these opinion pieces about the impacts of AI are AI-written (as is almost certainly the case here, a major trend in ACM and IEEE publications). Like, really? An LLM has a breadth of experience managing AI, and can recall it, and explain it to humans?" — a skeptical reader in the thread
No proof followed, just a shrug that ACM and IEEE opinion sections have run AI-flavored commentary about AI before. Nobody in the thread confirmed or denied it either way.
"LLMs made programming easier. Capitalism prevents workers from capturing that value."
The clearest rejection of the essay's core framing came from a developer with a decade-plus HN history and over 9,000 karma, arguing about who keeps the productivity rather than whether it exists.
"It's absolutely not true that it 'just' moved the difficulty around... the reality is I can get more done, and get better work done... than ever before. LLMs made programming easier. Capitalism prevents workers from capturing that value." — a developer with a decade-plus HN history and 9,000+ karma
A freelance engineer: the code is better, the job isn't fun anymore
A few replies down, someone made the opposite complaint.
"I don't think I write or read or code anymore... at this point AI writes better code than me and I'm not the average developer... Codebases have never been healthier, cleaner, better documented... and thoroughly tested as they are now... Needless to say I no longer enjoy the job anymore and thinking of changing domain." — a freelance product engineer focused on functional programming and effect systems
He isn't arguing the work got harder. He's saying it stopped being his.
Who inherits the AI patchwork later
The most confident dismissal of the whole debate came from a commenter whose bio tracks a monthly "probability of AI doom," sitting at 98.2 percent this month.
"It's way easier. People don't need to read code anymore... The way to think about coding agents is that a good one should in theory literally replace the developer entirely. No, a whole organisation of developers... AI is making programming an irrelevent field." — a commenter who tracks a monthly "probability of AI doom" in his bio
A DevOps engineer whose bio lists databases and Linux answered with a forecast instead of a feeling.
"If my experience with RDBMS generalizes, I predict that this trend will result in companies hiring skilled developers to unfuck their codebases... they will identify massive structural issues, propose huge lifts to redo it correctly, and will instead be told to do patchwork fixes to stop the bleeding... 'Our p99 on this endpoint is breaching' becomes 'I've added an index for this query'... when the actual problem may well be that you're storing huge blobs that bloat your working set, cause buffer pool churn, and saturate your NIC... Eventually you'll decide to shard, which adds complexity, when the actual solution would've eliminated all of those issues at once." — a DevOps engineer whose HN bio lists databases and Linux
Both readings can be true in the same codebase. The essay named judgment as the new hard part, but it never said who pays when the judgment is bad: the person who shipped fast has already moved to the next feature by the time someone else inherits the decision. That's the argument the thread actually had, underneath the one about recall and judgment: not whether programming changed shape, but who's still standing there holding it when it breaks.
Get breakdowns like this every day
The top Hacker News stories, unpacked — by email, Telegram, or RSS.
Jane Street's 'Incremental' is the trick Goldman called 'Node Purpling' 30 years ago
Jane Street put its internal OCaml library up on GitHub: code that recomputes only the parts of a system a change actually touched. In the thread, a developer who'd spent 13 years at Goldman Sachs said the bank ran the identical graph under that exact name three decades earlier.