German's central thesis is that the engagement economy has exhausted its easy reserves and is now 'fracking' for diminishing dopamine yields. He points to Duolingo's threatening owl, LinkedIn's manufactured pings, and Snapchat's resurrected streaks as evidence that each intervention produces less return, forcing platforms to dose harder rather than reconsider the underlying model.
The editorial argues that the diffuse 'capitalism did it' narrative obscures the fact that named individuals wrote, reviewed, and QA'd each dark pattern. Streak-resurrection modals don't fall from slide decks — they trace back to real GitHub usernames, which means moral agency is concrete and individual, not abstract.
Half the top comments named specific apps they had deleted in the last 90 days, treating aggressive notification and streak mechanics as the trigger. The lived experience in the thread is that the escalation has crossed a threshold where users now recognize manipulation and respond by leaving entirely.
A former engineer on a major social app's notification system confessed that every increase in notification aggressiveness was justified by short-term retention numbers, with PMs leaning on the metric to override discomfort. The implication is that the org chart and incentive structure manufacture complicity even when individual contributors privately object.
Igor German's essay 'Dopamine Fracking' hit #2 on Hacker News with 657 points and 400+ comments, framing the modern app economy with a single industrial metaphor: the easy reserves are gone, so platforms now pump high-pressure engagement tactics into thinner and thinner psychological seams. The piece walks through the standard catalogue — streaks, badges, red dots, variable-reward notifications, the chip-off-the-Skinner-box pull-to-refresh — and argues the diminishing returns are now visible in the product itself. Duolingo's owl threatens you. LinkedIn invents reasons to ping. Snapchat resurrects expired streaks. The yield per intervention is collapsing, so the interventions get louder.
The HN thread did the thing HN threads do: half the top comments named specific apps they'd deleted in the last 90 days. The other half were engineers and PMs admitting, with varying degrees of squirming, that they'd shipped exactly the patterns under discussion.
The essay's real argument isn't that these patterns exist — everyone knows they exist — it's that they're now obviously failing, and the industry's response is to dose harder rather than rethink the model.
There's a comfortable narrative in tech that dark patterns are a product problem, or a leadership problem, or a capitalism problem — anything except a problem with named individuals shipping named commits. The narrative survives because it's diffuse. The streak-resurrection feature in your photo-sharing app didn't fall out of a slide deck; somebody wrote `if (lastStreakBrokenWithinHours < 48) showRecoveryModal()`. Somebody else reviewed it. A third person QA'd it on a staging build. The chain of moral agency runs through real GitHub usernames.
The most upvoted reply in the HN thread came from an engineer who'd worked on a major social app's notification system. Their summary: every escalation in notification aggressiveness was justified by a short-term retention metric, every PM who pushed it got promoted, and every engineer who built it told themselves the next role would be different. The 'next role will be different' bargain is the single most reliable engine of dark-pattern accumulation in this industry, and it has been since at least 2015.
Compare this to what happened in adjacent disciplines. Civil engineers refuse to stamp drawings they consider unsafe; the refusal is enforceable, the liability is personal, and the professional culture treats the refusal as routine rather than heroic. Pharmacists refuse fills. Lawyers refuse cases. The software industry has spent two decades insisting it's too young, too fast-moving, or too economically dynamic to support equivalent norms. The dopamine-fracking essay lands now because the second half of that argument has stopped being persuasive — the patterns are too well documented, the harms too well studied, and the engineers too senior to plausibly claim they didn't know.
The interesting question isn't whether dark patterns are bad. It's why an engineer with seven years of experience, six competing job offers, and a fully vested first tranche of options still ships them.
The honest answer is usually some mix of: the specific feature didn't seem that bad in isolation; refusing one ticket wouldn't change anything; the PM would just find someone else; the engineer's stack-rank review is six weeks away; the alternative job market is fine but not infinite; and 'principled refusal' isn't a phrase that survives contact with a performance improvement plan. None of these are stupid reasons. They're also, collectively, why we are where we are.
If you're a senior engineer who actually wants to do something about this, the leverage points are smaller and more boring than the discourse suggests. They look like this:
When a notification-cadence ticket lands in your sprint, ask in the PR description what the expected delta in 30-day retention is, and what the expected delta in 'users who report the app as annoying in the post-uninstall survey' is. If product can't answer the second question because nobody runs that survey, that's the artifact. Make the missing data visible in writing, in a place that gets archived, and you've shifted the conversation from 'engineer refusing work' to 'engineer asking for completeness.' This is a meaningfully different political fight, and it's one you can win.
For staff-and-above engineers: the design-doc template at your company probably has sections for scalability, observability, and rollback. It almost certainly doesn't have a section for 'mechanism of behavior change' or 'predicted user-reported friction.' Add them. Doc templates are one of the highest-leverage cultural artifacts in any engineering org, and they're usually owned by whoever cared enough to send the PR. Be that person.
For IC engineers without that political capital: keep a private list of the patterns you've shipped that you wouldn't ship again. Reread it before your next job search. The list is the only honest input into the 'next role will be different' calculation, and most engineers never write it down because writing it down makes the bargain too visible.
The fracking metaphor is good because it's load-bearing in both directions. Yes, the easy psychological reserves are tapped out. But fracking also has a regulatory endgame — eventually the externalities get priced, the practice gets restricted, and the operators who built their whole business around it get repriced too. The EU's Digital Services Act is already pricing some of this. The US won't move first, but it doesn't have to. The more interesting timeline isn't regulatory; it's labor. There are now enough senior engineers in the market who have personally shipped patterns they regret that 'we don't do dark patterns here' is becoming a recruiting filter rather than a press-release line. The apps that survive the next five years will be the ones that figured out, before they had to, that the engineer who refuses the ticket is cheaper than the regulator who writes the fine.
Reminds me of Adornos "Dialektik der Aufklärung" and its take on what he calls the "Kulturindustrie". Almost 100 years ago he foresaw how the cultural offerings of society get commodified and chopped into bite sized chunks for each individual to receive theirs. He did not forsee
The strawberry example reminds me of the Instant Mashed Potatoes non-book review [0].> Since World War II and the large-scale industrialization it fully unleashed, a core method driving ‘progress’ across many different fields of human endeavor has been to shred something real and reconstitute it
As someone under 40 who never had any social media, I cannot overstate the negative impact it's had on my peers and their behaviours. Worst thing to ever happen to society imo, I feel for the younger ones who grew up with it.
I've found that this engineered optimization has a more pernicious side effect: killing curiosity.Lack of complexity stunts the desire to become curious - to give reasons to look closer, ask questions, compare experiences - and ultimately develop 'taste'.When everything is optimized i
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.
I love this term - I think it beautifully describes the direction that at least, YouTube is heading towards. Take for example, this racket where a channel copies popular (non-kids) creators’ parody work, splits the screen in half with the content on left, adds a completely random DIY type video on t