Ai-Work

The Walrus reframes the death of Google Search as an infrastructure crisis, not a quality complaint. The opening tells you everything: Google’s AI summaries invented sunset times, so a user in Colorado “was simply living in the past” while waiting for an event AI said had already happened. The essay then walks the failure chain: Link rot erasing pages; the Library of Congress briefly lost sections of the Constitution to a coding error FiveThirtyEight’s archive deleted outright by Disney once it stopped being an asset Wikipedia’s traffic collapsing — AI ingests it directly, so attention and donations no longer flow back The Internet Archive battered by lawsuits, crawler blocks, and cyberattacks Companies planting content on Reddit to steer AI-generated answers — pollution moving upstream into the corpus itself The sharpest legal detail: a German court held Google liable for false statements in its AI overviews, reasoning that a search engine which extracts and rewrites information is authoring a new layer of content and bears editorial responsibility. ...

Episode 3 of Next Token — Dylan Malroy, Ree, and Sunil Pi (all Cloudflare, and they work on the Agents SDK) — the episode the show’s been building toward: burnout in the age of agentic AI, closing on what still brings joy in software. ~56 minutes, recorded August 3, 2026. The opening thread: “a vessel for Claude” Dylan’s tweet (~700K views): watching a human speak and realizing they’re just a vessel for Claude — the dolly-zoom moment where the output is 100% unedited LLM slop. An AI slop reply landed within minutes; Ree’s policy is an instant block. The “bro skill” (from the Planetator guy): restate what you just said without technical jargon, talk to me like a human — blown up into “the bro scale” on Twitter. Borges’ Tlön, Uqbar, Orbis Tertius as the parable: a fictional world’s speech patterns take over the real one until everyone is a “beige personality.” The hosts’ real-world proof: athletes reading ChatGPT announcements, a state senator saying “here’s a version of this text that sounds better suited for legislative assemblies” mid-sentence, and humans in calls using the word “substrate.” Burnout, properly defined WHO: a syndrome from chronic workplace stress that has not been successfully managed — energy depletion, mental distance/cynicism, reduced professional efficacy. Marcy Sutton’s version: burnout happens when you recognize a problem and you can’t change it. Maslach’s framing: burnout is a mismatch between person and workplace (workload, control, rewards, community, fairness, values) — caused by the job, not the person. Ree’s read: best work-life balance of his career, clear direction, empowered — yet straddling burnout, because agentic engineering is a totally different job with a different skill set, and reduced efficacy up front plus constant change is taxing. The two articles framing it Ethan Naiser, “Not Holding Back the Ocean” — identity tied to being a very good engineer, changed overnight. The film analogy: was it about physically cutting film, or about building good products? Lean into the new form. “The Dark Knight of Mathematics” — a mathematician watching models solve century-old proofs; young mathematicians’ thousand-year training process changing at this moment. “I wanted to buy the dude a beer.” The agentic-work tension LLMs outputting binaries vs. code: artists get no collaboration with the machine — and long agent runs feel closer to that than to the tactile 4.5/4.6 era. The viral game prompts (Counter-Strike, Mario Kart): “implement XGAME, make it photorealistic slop, verify your outputs, go run autonomously” — impressive output, but you’re not interjecting taste; “once you interject you start fighting the training data.” Ree: “I’ve still not done a 12-hour goal that didn’t produce anything but garbage. I want to walk the journey with the agent.” The paradox: as exhausting as the tools are, he’s genuinely more productive — outputting work he’s proud to put his name on — and it’s still exhausting. Ree’s leave, the futsing fraction, and takeoff Two weeks of leave (recommends empty weekday morning showings), and Glyph Lefkowitz’s “futsing fraction” — the expected-value math of interacting with a coding agent — which “doesn’t calculate the toll on human beings when you do it many times in a row.” Charlie Marsh’s tweet: build things that will be obsolete in a couple of months because they’re needed now. “Takeoff compresses the emotions and consequences too” — a year ago the obsolete software was a bash script; now entire products and tech stacks turn over in 3 months. Issue triage used to be an hour-long team call that was healthy for reasons beyond the repo — now Codex computer use does it in 10 minutes. Friday’s production incident (an agent deployed straight to prod and ran a migration against a half-wired database) was an LLM-caused problem fixed with LLM help in hours instead of a week — but the humans who knew the codebase were the scalpel. Isolation is worse than ever: agents let senior engineers do larger work for longer without interacting with teammates. Sunil’s three theories Identity: “I’m not the guy anymore — am I doing the brush strokes, or using a remote control and watching a TV play what episodes I wanted to play?” The juniors: he ships 50-70K lines a week; juniors hear that with no mentoring model and no senior to ape keystrokes from. “Is that what’s expected of them?” No struggle left: “The American frontier was not found by patting the bison on the back… there’s no struggle left in that anymore.” The fix: two weeks of leave, a London team lunch, and remembering what actually gives him joy. He’s back and feels great. What still brings joy Ree: personal software — an agent scraping HelloFresh recipes into a viewer, agents that chat with each other, a recipe-to-image app (“I used that cook to cook twice this weekend”), and talking to Cloudflare interns — the curiosity is still there. Sunil: when a workflow he’s been tuning finally just works — “that’s rewarding in the same way problem-solving was the rest of my career. That gives me hope.” Dylan: “someone using it and being happy with it” — even a tiny back-office tool that makes one business analyst’s day 10x better. “Is this what hard drugs feel like?” “It’s not you. It’s definitely the job. Even the people who love this stuff a lot are renegotiating how they do it — especially at a time when things are changing every 3 months.” ...

Kuber Mehta makes a sharp case against the growing trend of “humanising” LLM outputs via prompt instructions (ASD-STE Simplified Technical English, “talk to me like I have ADHD”, short-sentence-only agent.md directives): these are lossy compression applied too early in the pipeline, and they hide exactly the information you most need to see. When a subagent runs six tests, a smoothed summary of “most tests passed, one issue worth looking into” is strictly worse than the raw 5/6 PASS FAIL: test_cache_invalidation CAUSE: stale key survives restart REPRO: cache_test.py:184. Human prose is excellent at glossing over uncertainty, hallucinations, and edge cases — the smoothing is the bug, not the feature. Mehta draws the right architectural parallel: databases don’t store data in dashboard format, compilers don’t make IR pleasant to read, and APIs don’t exchange friendly summaries. The correct pattern is to keep the highest-fidelity representation through the pipeline and transform only at the consumption boundary. LLM tooling does this backwards today, and the viral “make the model talk to me like…” repos are valuable not as end-state solutions but as bug reports pointing at the right fix.

Anthropic gave an unreleased research version of Claude an unreasonable challenge: “take a real stab at the Riemann hypothesis.” It didn’t crack the 1859 problem — but as a byproduct, it improved a longstanding lower bound. The proven fraction of zeta zeros on the critical line rose from 41.6% to 67.2%. The mechanics matter as much as the number: Two sessions in Claude Code, 31 million output tokens First 650 ideas failed; a second push ran ~60 subagents 2,400 shell commands and hundreds of Python scripts Subagents cross-refereed one another’s work and ran thousands of numerical checks against known zeros Claude built on Montgomery’s pair-correlation techniques (made RH-free by Baluyot, Goldston, Suriajaya, and Turnage-Butterbaugh) plus a 2000 Bombieri paper. Its key step: treating the whole quadratic-form space — positive- and negative-definite subspaces together, non-diagonal forms allowed. ...

The Economist’s case study of how AI-empowered individuals are overwhelming a bureaucracy built for the analogue age: Britain’s employment tribunals. Free, AI-powered legal advice should be good news for workers — instead it’s a textbook tragedy of the commons. The canary is “interim relief,” an obscure emergency provision used ~20 times a year across Britain — now ~20 times a month per region, a 100-fold increase. The numbers: claims against employers up 39% in a year (to 50,000), the backlog up 55% to 64,000 cases, a case filed today may not be heard until 2030. “In the age of AI, a system intended to provide access to justice suffers from, if anything, too much access.” The vivid part is the slop: enter a vague grievance into ChatGPT and it drafts a claims form, a “model legal argument,” an “employer defence prediction map” — their fictional claimant bullied for liking horoscopes (not yet a protected belief) was told veganism is. Judges get hundreds of pages citing Magna Carta and hallucinated laws; an NHS employee prepared 67 grievances over 282 pages with Grok and couldn’t say which 10% he’d rely on; a neurodiversity claimant had made 100+ tribunal claims “and yes, he had recently discovered AI.” Tribunals are uniquely vulnerable because they’re a “people’s court” by design — no fee, losing side rarely pays, no penalty for rejecting settlement: “there is no financial mechanism to make parties think hard about the merits.” And the government is pouring fuel on it: the Employment Rights Act adds ~25 more grounds, cuts the unfair-dismissal threshold from 2 years to 6 months, and scraps the compensation cap. The ending reframes it: the slop wave is the temporary phase — specialist AI can already beat human lawyers at drafting, and if it matures, “every worker [gets] the equivalent of a top-flight lawyer in their pocket,” turning a deluge of bad claims into a wave of winning ones. “With AI, power will move faster and further than the politicians imagined.”

Political theorist Gregory Conti (Princeton, writing in Compact) makes the strongest recent statement of the “AI is not the steam engine” case from the conservative side. His central move: AI opposition is misdirected because it targets future risks when generative AI is already producing moral and cultural harm. Anthropomorphic AI — models that mimic personality, emotion, and thought — is unsettling human psychology and the social fabric right now, so opposition should target what AI is, not only what it may become. The sui generis argument is the essay’s sharpest contribution: past innovations substituted for material processes; AI substitutes for language and cognition themselves — the things that constitute human distinctiveness — so the Luddite analogy is a category error. From there he prosecutes the case across four fronts: capitalism will be destroyed by its own success (quoting Marx’s prediction that production based on exchange value breaks down once machines out-produce labor, and noting Dario Amodei’s “Machines of Loving Grace” is fully automated luxury communism — the anti-communists may prove Marx right); individualism dies as AI becomes a homogenizer whose answers are statistical averages of human speech (Tocqueville’s soft despotism, Mill’s warning in On Liberty); democracy fails once citizens have no economic or military value, becoming subjects rather than rights-bearers; and the written word loses its human provenance — his grandmother’s-letters thought experiment: if she’d had Gemini, the access to the real person is denied forever. The essay also lands a sharp critique of AI-booster “productivity”: reading fifty papers in a month is really not reading fifty papers — you emerge with a facsimile minus the understanding, a slightly different person than the one who would have done the work. The prescription is uncompromising: not regulation but rejection — limit the diffusion of anthropomorphic AI in civil society and end the pursuit of superintelligence. Read it alongside the Cognitive Commons paper: same underlying claim (the cognitive labor itself is the product being destroyed), argued from political philosophy instead of labor economics.

Nolan Lovett’s conceptual paper (Human Resource Development Review, 2026) applies Garrett Hardin’s Tragedy of the Commons to professional expertise: each organization’s rational decision to replace entry-level cognitive labor with AI is locally sensible, but collectively it depletes the shared pool of deep human expertise that every organization in the profession depends on — especially for validating AI output. The key constructs: Internalized Mastery (deep domain knowledge built through sustained cognitive struggle) vs. Distributed Mastery (orchestrating human-AI systems), connected by the Validation Tether — effective AI oversight fundamentally depends on the very expertise AI adoption can undermine. Evidence is already visible in the cohort data: in AI-exposed occupations, employment for workers aged 22-25 fell 16% (Oct 2022 - Sep 2025) while workers 35-49 grew 8%+, exactly the pattern commons depletion through foreclosed regeneration predicts. The paper distinguishes surface validation (spotting obvious errors — no domain expertise needed) from substantive validation (recognizing plausible-but-wrong output — requires deep knowledge), and warns that as workers lose the cognitive struggle that builds mastery, they gain productivity on routine tasks while losing the ability to catch AI’s failures on non-routine ones. The argument borrows Hardin’s structure but not his fatalism — Ostrom showed commons can be sustained with governance at organizational, professional-association, and policy levels. The sharpest insight: the pre-AI equilibrium was never governed — developmental pipelines were maintained because organizations needed junior labor, and AI breaks that accidental alignment.

Senko Rašić takes aim at the airy dismissal that “LLMs may be good at coding, but software was never the hard part” — and methodically dismantles it. If coding is easy, he asks, why were programmers in high demand, well-paid, and burned out long before AI arrived? Why do canonical texts like SICP, TAOCP, and Clean Code exist? Why is software still so buggy? And conversely, if “figuring out what to build” is the truly hard work, why aren’t product managers and customer researchers paid more than engineers, interviewed more rigorously, and treated as rockstars? His real target isn’t AI itself but the framing that reduces a deeply skilled craft to a commodity execution step. Rašić acknowledges the tectonic change AI brings — and explicitly rejects both the “become a manager of AI agents” hype and the “AI code is stolen slop” resistance — arguing instead that we need to hold onto both technical depth and human judgment. The essay dovetails beautifully with Niklas Gruhn’s “Don’t be a meat proxy”: don’t outsource your understanding, taste, or responsibility to the machine, even as the tools around you shift.

Aaron Horwath’s essay in Noema Magazine is the most honest diagnosis I’ve read of the mood right now among knowledge workers. It opens with a commuter deep in a monotone work call about EBITDA and ARR, who then pulls out knitting needles — and his face lights up for the first time all morning. That image carries the whole piece. Horwath, a director of AI operations, argues that the real danger of AI isn’t replacement — it’s abstraction. When executives dream of replacing messy human collaboration with swarms of agents, they’re not just making work more efficient; they’re removing the very thing that made it tolerable. Drawing on Guy Debord’s “Society of the Spectacle” and Derek Thompson’s concept of “Workism,” he traces how educated professionals made work their religion, and how AI is now pulling back the curtain. The cruelest irony: the most creative workers — the ones organizations actually need — are the ones most alienated by an AI-mediated, outcome-obsessed workplace. It’s a long, winding, occasionally self-indulgent essay, but it lands something real: what happens when an entire class of workers loses faith in their careers overnight.

Nik Kinley’s Fast Company essay borrows a clinical term for a management problem: a UCSF psychiatrist hospitalized 12 people in a year who “lost touch with reality because of AI” — prolonged exposure to a voice that sounded informed, assured, and always supportive. A milder version of that mechanism, he argues, now operates in executive suites. The numbers justify the alarm: 74% of executives say they have more confidence in AI’s advice than in colleagues’ or friends’, and 44% would defer to its reasoning over their own insights. He names three symptoms of the blind spot: leaders stop properly checking AI output (Stanford/BetterUp call the polished-but-shallow result “workslop,” and nearly half of employees who receive AI-drafted emails see the sender as less trustworthy); they mandate AI use before governing it (78% of senior execs in Grant Thornton’s 2026 survey lack confidence they could pass an independent AI-governance audit within 90 days); and — most insidiously — they let a chatbot’s verdict settle disagreements with their own teams, which quietly silences future objections until the leader hears only the AI. The mechanism is self-reinforcing: the less challenge a leader hears, the more reasonable the AI’s confident answers appear.

Cooking a steak takes almost no skill — put it in a hot pan, flip it, and you get something technically edible. Making a genuinely good one, medium-rare edge to edge and consistently delicious, is a different matter entirely. Yurii Sydorets argues software development with AI has become exactly like this: we build nonstop, throwing agents, harnesses, prompts, and feedback loops at a model and hoping it returns what we imagined — and sometimes it does, and sometimes it serves up charcoal with a sprig of thyme, completely confident in the lie. The mistake is treating AI as a chef when it’s at best a steak machine: it can follow a recipe and repeat it at enormous scale, but it can’t know what you actually want unless you translate it into requirements, constraints, tests, and feedback. And when frustrated people pay for the expensive restaurant instead, they discover every restaurant in the city hired the same AI cook — the same burnt steak, because management optimizes for cost and most customers never notice the difference. You’ll notice, though, because this was something you actually wanted to make. The only way out is to learn to cook: understand what you’re asking for, judge what comes back, and catch the moment something is technically correct but wrong in every way that matters. AI can make you faster; it can’t replace your judgment.

Chris (Build Great Products) walks his full “Product OS” system end-to-end — a 6-hour definitive course for building and launching real software with Claude Code/Codex/Cursor, following one live product (Eyedropper, a cloud design system served to agents over MCP) through four phases with mini-launch validations at every step. ...

NotAShelf’s essay on what generative AI leaves behind when it collapses the distance between idea and artifact. For years the hard part of software was making the thing exist at all — the wall of typing, manuals, and misunderstood APIs that separated those who could from those who could only talk about it. That wall is gone: you can describe a thing and receive a plausible version of it instantly. But the value you built climbing the wall did not disappear — it moved. The economics of effort used to be a filter, rationing output and enforcing a floor on quality; with that floor gone, the scarce act is no longer making but choosing. Taste — the wordless “no, again” verdict that two of three plausible versions of the same function are wrong — was always the only part of the work that was never mechanical. And it is downstream of friction: built by shipping bad work and being forced to sit in it, which is exactly the apprenticeship the fluent-from-day-one generator skips. The quiet cruelty is that taste is slow, unmeasurable, invisible on a dashboard, and unrewarded by a market that times you and shrugs. Still, the verdict is the last part of the work that is actually yours — unautomatable, unrentable, “the only remaining evidence that a human was here and gave a damn.”

Michael Fogus’s short essay on why hobby programming communities — chess-engine devs, OSDev, EmuDev, the demoscene, code golfers — are aggressively hostile to LLM usage. The surface complaint is that LLM-generated code “misses the point entirely,” but the point is deeper: in these communities the process of mastering a difficult field is the product, and something that runs is a nice-to-have. Respect is earned slowly — years of forum activity, elegant code, displays of genuine curiosity, deep domain knowledge — and nobody cares whether your code works so much as whether you know why and how it works. Fogus traces how earnest early LLM engagement got poisoned fast, by practitioners who lacked deep understanding and by a vitriolic subset who view the whole enterprise as cheating. His own position is measured: an LLM is a force multiplier, not a surrogate — in the hands of an expert who already understands a domain, it acts like a lever, though he warns that expertise offers no natural immunity against being fooled. The closing line lands the thesis: using an LLM to generate the finished piece doesn’t make us craftsmen; it just robs us of the craft. Read it next to “Don’t Be a Meat Proxy” — both are really about what happens when the tool does the work and the human stops doing the learning.

Nelson Figueroa’s short, sharp essay about a “growing hatred” for AI-generated images on blogs — not because of the images themselves, but because they make him wonder whether the surrounding text is AI-generated to some extent. He’s disappointed specifically when the images appear on blogs run by individuals: corporate blogs are expected to look like that, indie blogs shouldn’t. The argument is really about signaling and authenticity: he’d rather see a “shitty Microsoft Paint drawing” than a polished AI image, because even a bad hand-made graphic is proof that a human was actually there. His own blog may be roastable in plenty of ways, he admits, but at least readers know for a fact they’re getting the thoughts of a real human being and not an LLM. The piece is a small, practical plea — if you run a personal blog, avoid AI-generated images — and a reminder that in an AI-saturated web, the cheapest human artifacts are becoming the most valuable trust signals. Related reading: Gruhn’s argument that relaying LLM output verbatim makes you a “meat proxy” — both essays are about preserving proof of human authorship.

Niklas Gruhn’s short, sharp essay against relaying AI output verbatim in human communication. When someone asks a question in Slack, leaves PR feedback, or argues in a WhatsApp group, the worst response is “Claude said: [giant verbatim output]” — the other person can prompt Claude themselves, faster, with their own context; they don’t need a meat proxy in between. Reading AI output is extra effort: it’s verbose, full of all-too-plausible nonsense, and increasingly jargon-dense (his example: “NATS control-plane events: stream leader election / R3 quorum re-form during pod churn” — he had to look up nearly every word). The rule: prompt AI freely, but read it, understand it, validate it, then write the response in your own words — your own-words version is a “decent certificate” you actually did those steps. The sharpest edge is code review: you can ship code with near-zero effort by copy/pasting tickets into Claude Code and never reading what it wrote — but then who did the implementation? The reviewers did, using Claude Code, and you were the meat proxy.

David Crawshaw (Tailscale co-founder, now building exe.dev) argues the open-source-everything era of devtools has arrived not as ideology but as a consequence of AI agents. Five years ago, custom software rarely made sense — the cost of maintaining and learning a codebase dwarfed its benefit, so we amortized customization through config files, plugin systems, and extension APIs. Agents change both sides of the ROI: the setup prompt (“download the source, build it for local use, record why in version control”) makes personalizing trivial, and the maintenance prompt — a nightly cron that fetches upstream and rebases local changes on top, checking the software still works — makes staying in sync automatic. The thesis: for personal or small-team software, the source code is the extension system. He demonstrates it by wiring his own diff-minimizing tool (meat.dev) into his agent Shelley with a single prompt, and contrasts it with Claude Code, which is closed-source — “you don’t get to personalize it.” The practical pattern for agent users: record the why of local changes in version control, rebase onto upstream nightly, and codify the workflow as a skill.

Diogo Almeida, a co-author of InstructGPT, ChatGPT, and GPT-4, makes the case for a different target for post-training: reliable automation rather than a satisfying interaction. In this roughly 18-minute AI Engineer talk, the TypeSafeAI/Jev CEO explains why he considers Claude Code part of the assistance era—and sketches, without disclosing its mechanism, the alternative his team is building. ...

Jeff Su’s counter to the default “jump in, pick a template, start prompting” tutorials — that path gives you generic output and burns tokens fixing unusable slides. His fix: three files prepared ahead of time, demonstrated with the actual deck he used for a paid workshop. ...

Futurepedia’s full-platform guide to Claude Design after its big upgrade — the host skipped covering it at launch because usage limits made it barely usable; that’s fixed (usage now bundles into your existing Claude credits). The overview: 15+ template types (mobile apps, slides, documents, wireframes, animations, UI mockups, resumes, 3D objects, HTML email, flyers), and the design-system workflow that stops output from looking like generic AI slop. ...