On August 18, 2026, a programmer going by ramesh31 posted on Hacker News with a question for a title: *Does anyone else feel like everything is pointless now?*
He laid out his inner ledger: New software releases? Who cares. Learning new CS concepts? Whatever. Prepping for interviews? Hard to say the job will even exist in a year. Side projects? Just another shovel of dirt on the AI slop pile — besides, the guy bagging groceries at the supermarket could do it with AI too. Everything is "done and cooked with AI" now.
He ended with an analogy the whole thread kept chewing on:
> "It's like being given cheat codes in a game. You can do everything, yet it all feels meaningless. Our purpose and joy in life has been taken away entirely."
The post got 297 upvotes and 175 comments. The day before, another thread had asked whether this is the worst time in history to be a human, and people passed the same cheat-code metaphor between the two. This isn't one person's 3 a.m. breakdown — it's a batch of people's shared mood. One darkly funny detail: after hitting 116 points, the post was flagged off the front page, yet the comment section burned all day. Some laughed bitterly — guess it really hit a nerve.
Scrolling back through ramesh31's posting history reveals a complete anxiety trajectory: in 2022 he asked whether self-taught developers all fear interviews; in February 2025 he asked what interviews look like now that everyone uses AI — that post got 535 upvotes. Over four years, his question drifted from "am I good enough" to "does this thing even exist anymore."
I read all 175 comments. My takeaway: there was no debate here — only the reflections of three kinds of lives in the same well.
I. The Person Playing Chess with an Engine
Commenter bah9 wrote a highly upvoted passage:
> "Programming for me was chess, and AI is a chess engine. Of course I care about winning — shipping the product — but the game is being played by the engine, so winning brings no joy. Strangely many people don't get this. Apparently 90% of programmers hate chess; they only love winning. The sad part: now every 'grocery bagger' can win too — so what exactly is your value?"
Hidden in there is a distinction he may not have noticed: playing chess and winning are two different things. The engine killed the latter. Someone followed up: worst of all, now everyone fancies themselves Carlsen.
Another commenter, frankie_t, twisted the cheat-code metaphor in a different direction. The problem isn't "I got the cheat code," he said — it's that "someone who cheats entered the game," so the game itself no longer deserves to be played seriously. Two different pains: one is my ability being diluted, the other is the rules being broken. Eerily, both lead to the same emptiness.
II. The Person Who Finally Finished the Project
thatmiki, a 40-year-old front-end engineer, showed up too. He said he completely understood the OP — then pivoted:
> "On personal projects — I'm actually finishing things! The joy of creating hasn't vanished; it's changed. I miss the puzzle-solving moments too, but watching my ideas become real, useful, and polished is deeply satisfying."
He added a line more informative than all the complaints in the thread combined: without his twenty years of front-end craft and UX judgment, these projects would be "way more slop."
throw-the-towel told a more vivid story. One night he stayed up until 3:30 a.m. hand-writing a CLI tool that converts shapefiles to vector tiles; two nights in, it still wasn't done. Opus could do it in about five minutes. His closing line:
> "The robot may be better at eating cheese than me, but that doesn't stop me from eating cheese."
III. The People Saying "Reality Has No Cheat Codes"
chr15m dismantled the metaphor directly: "Life has no cheat codes. Physics still applies, and reality has an astonishing number of details." The moment you actually build things, "anything is possible" falls apart — wherever the constraints and the frontier are, that's where the boundaries are, and that's where the fun is.
Someone offered a work example: have an agent touch genuinely high-stakes ops — a TLS termination migration with coordinated DNS changes for thousands of customers, many Fortune 500 — "it breaks after the first context compaction. Every time."
The requirements remain, the incidents remain, the edge cases remain. They've never sympathized with anyone.
The most striking comment came from Balgair, who claimed to be the very "grocery bagger" from the OP:
> "I'm more like the bagger in the post. AI, to me, means finally getting tools. The world has opened up now. I can build websites; I can do the things I always wanted to do with a computer. Your world is finally open to me."
The same fact, reflecting two different lives. This sentence deserves to be pasted at the end of every AI-anxiety analysis.
Psychology Answered This Long Ago
The real puzzle: what exactly is that "post-cheat-code emptiness"?
Psychology delivered its verdict in 2018. In that year's review of the *paradox of effort*, Michael Inzlicht of the University of Toronto and colleagues wrote a line that reads like it was addressed to ramesh31 eight years in advance:
> "When zero-effort alternatives are readily available and obvious, exerting effort itself feels meaningless."
The brutal part: the devaluation doesn't require you to use the zero-effort option. Its mere existence dilutes your effort. That explains frankie_t's "the opponent cheats," and why many programmers who barely use AI feel hollowed out all the same.
There are earlier experiments too. In 2004, Kruger et al. ran a classic study: the same poem, described to one group as taking eighteen hours to write and another as four hours, was rated significantly higher in quality and value by the first group. Nothing changed except the labor behind it (honestly noted: the 2023 replication had mixed results).
The IKEA-effect experiments are blunter. Have people fold origami themselves, then bid on their own work: about 23 cents on average. Bystanders looking at the same crooked frogs offered only 5 cents. The self-builders priced amateurs' work close to what bystanders paid for experts' work — without realizing how much they were overvaluing it. And a cold tail to the experiment: when the work was assembled and then immediately dismantled, the effect vanished. Labor must be completed and must persist for love to attach to it.
Put together, 2026 makes sense. AI extracted "effort" from most software labor: the folder is still at the desk, but the crane was folded by someone else, and another can be folded in three seconds. You stare at it and can't love it — and per the psychology above, that love was never supposed to appear.
The game scholar Jesper Juul argued in *The Art of Failure* that games are the only art form that deliberately makes us fail. Failure supplies the stakes; only then does winning carry weight. Remove the possibility of losing, and winning means nothing.
But hold off on conclusions. A 2020 CHI PLAY study specifically examined player cheating, with deflating results: for casual players, moderate cheating made the experience *better* — more relaxed, more in control. "Cheating ruins the game" held mainly for achievement-oriented players.
That deflating study may be the key to the entire thread. It precisely predicted the split in those 175 comments: those who code to *win* were pierced by AI; those who code to *play* were barely touched.
Chess's Thirty-Year Natural Experiment
So what's happening to this industry? Chess happened to run a thirty-year natural experiment, and the results were stranger than anyone predicted.
On May 11, 1997, Deep Blue beat Garry Kasparov 3.5–2.5; in the sixth game he resigned after only nineteen moves. Newsweek's cover called it "The Last Stand of the Brain." Kasparov accused IBM of cheating, demanded the machine's decision logs — refused; demanded a rematch — refused. IBM promptly retired Deep Blue. The human world champion loses to the machine, the machine exits the stage — by that script, chess today should be a museum exhibit.
Reality: chess.com reached roughly 250 million members in early 2026; about 1.64 million players hold FIDE ratings worldwide; the 2024 World Championship drew 12.9 million hours of total viewing; and 18-year-old Indian prodigy Gukesh became the youngest world champion in history. A game pronounced dead thirty years ago is now in its most flourishing state ever.
Go tells a different numeric story. When Lee Sedol retired in 2019, he told reporters perhaps the saddest quote ever from a player:
> "Even if I become the number one, there is an entity that cannot be defeated."
He was 36 that year, losing his farewell match 1–2 to a Korean AI — his one win came from a trick move exploiting AI weaknesses. Meanwhile, Korean Go's population math: overall numbers fell only ~4% over a decade, yet registered professional and insei players halved from 3,260 to 1,633, and university Go departments faced shutdown in 2022. The amateurs held steady; the professional pyramid contracted.
The engine didn't kill the people who play. It tore apart "playing chess" and "becoming elite through chess."
Kasparov himself left the experiment's conclusion in the New York Review of Books in 2010:
> "Weak human + machine + better process was superior to a strong computer alone and, more remarkably, superior to strong human + machine + inferior process."
This has a famous empirical backing. The 2005 first freestyle chess tournament allowed human-machine teams; the champions were two American amateurs rated 1685 and 1398, running three ordinary computers, who defeated the team behind supercomputer Hydra and a string of grandmaster pairings. Top-tier technology lost to humans with better process.
For programmers, the translation is effortless: memorizing opening books and calculating variations is the equivalent of writing CRUD. Once the engine devalued those to zero, what remains valuable is understanding, taste, process, and accountability. The agent that exploded during the TLS migration never lacked coding ability — it lacked "whose neck is on the line if this goes wrong."
The data also sides with "rewriting." The 2025 Stack Overflow survey: 46% of developers distrust AI output accuracy, exceeding the 33% who trust it; on the biggest pain point question, two-thirds chose "almost right, but not quite." METR ran a randomized controlled trial: 16 senior open-source maintainers handling real issues with AI were actually 19% *slower* — while estimating they were 20% faster. GitClear analyzed hundreds of millions of lines and found that in 2024, copied-pasted lines exceeded refactored-moved lines for the first time in history. Yet over the same period, GitHub developers grew to 180 million, and TypeScript topped language rankings for the first time thanks to the agent era.
Trust is collapsing, quality is degrading, and the population of practitioners and total innovation are both rising. Both things are true at once. The value chain is being rewritten, not erased — and during the rewrite, the previous generation's moats get filled in. For those defending the castle, that's a disaster; no need to dance around it.
Where Did the Joy Go?
Lee Sedol and Magnus Carlsen are the control group within one industry. Lee Sedol's coordinate system was "the summit of humanity"; the engine deleted that coordinate system, and his motivational system collapsed with it. Carlsen said on a 2022 podcast that he personally avoids engine analysis — "imperfect knowledge" is worse than no knowledge — "I love chess, I love to win." He chose to stay inside imperfection; that's the game he loves.
What got pierced was *winning* — it was never *playing*.
Psychology has a kindergarten experiment for this too, from 1973. Children who loved drawing were split up; one group was told they'd get a "good player" certificate for drawing. The certificate winners drew noticeably less during subsequent free play. The reward quietly rewrote "I like drawing" into "I draw for rewards." Once the reward vanished — or became something everyone got — the motivation had already changed owners.
AI did the same thing to a generation of programmers, except it withdrew the external rewards all at once: salary premium, scarcity, professional identity, the elite aura of "being able to code." What remains after the withdrawal — whether you still write code — is the real answer to the original question.
The thread's answers varied. preg_match is writing a new programming language, openly projecting "exactly zero" users besides himself in the AI era: "But I love languages, I love language design. Seeing my own language run and thinking 'I built this' — no joy can replace that." An unemployed commenter turned to photographing spiders, aiming to collect all three species of a genus in his state — achieved, now hunting the next genus. On chemo leave, someone writes code for himself and realizes what he always hated was the corporate environment; the love was there all along.
The most piercing, most stable line of the whole thread carries a Stoic flavor:
> "If your life's meaning can be taken away from you, it was always overly dependent on external factors."
That line can't silence everyone. Someone pushed back hard: intrinsic motivation doesn't pay rent — I could lose my house. Others pointed out this crisis sits atop a bigger backdrop — financialization, hitting shareholder numbers, manufacturing the appearance of value rather than value itself. All true. Don't paper over structural problems with "adjust your mindset." The work crisis is structural, and so is the pain.
But structure aside, each individual person still faces a multiple-choice question. It reads roughly like this: suppose the cheat code really exists and is already sitting in your pocket — do you use it, or treat it as the answer key?
The 40-year-old front-end engineer chose to finish first and ask later. The one writing CLI tools at 3:30 a.m. chose to keep eating cheese. And 250 million chess players, knowing full well the engine is unbeatable, are playing a game tonight anyway.
The game isn't over. It's finally just asked you a question: are you here to win, or here to play?