Full Width 
Shortcuts 
Sign Up 
Log In 
Articles 
Reading List More from matt.sh
agent-shell: stop spilling your outputs since the beginning of time, ai agents have wanted one thing at one thing only: redirecting stderr to stdout then piping the output to head or tail or grep or sed or awk. That’s all they really want. Except, they do it wrong the majority of the time (because “ai labs” are fundamentally incompetent at software engineering, as we’ve previously discussed before). Common ones I see routinely which are painful and require manual “reminders” into the agent to stop: uv run my-big-honking-test 2>&1 | grep -E "failed:|error:" | head -5 but then the test fails with a non-zero error code, the grep reports an error count, so then the agent RE-RUNS THE ENTIRE TEST with a different grep failure to “find the failing test case” except, the failing test case grep requires “guessing the future” regex, but the first attempt at “guessing the error regex” fails, then code agent continues re-running a 20 minute test suite with different grep filters until you remind it: “FILES EXIST! WE HAVE A FILE SYSTEM! WRITE OUTPUT TO FILES THEN READ THE FILES TO SEE THE ERRORS!” also, since current ai code agents have no feeling for “time” or “effort” they don’t care about running a 30 minute program-stderr-to-stdout-to-pipe-grep 10 times in a row even if each attempt takes 30 minutes. and a more toxic approach is where models decide to use live public URLs for reference material then curl | grep except, again, models try to “guess grep filters”, fail, then re-run the curl a dozen times with different “guess the regex” filters instead of saving the content locally and then searching through it (I often see claude get rate limited or blocked by upstream services because it “curls and greps” the same URLs in a loop, and you have to remind it every time “hey, files exist? you know? use files?”) I’ve run some bulk analysis about how often models “do the wrong thing” or “waste time guessing” (instead of following more proof-based data practices) over my claude session history (you do know all your claude session and all commands and tool calls are saved at ~/.claude right?) and here’s results: ways models try to read things Pattern Count Share of Bash commands Piped into a filter (grep/head/tail/awk/sed/etc.) 123,637 62.8% Bare narrowing command, no pipe 22,963 11.7% – Any pipe at all 135,413 68.8% Bare full-file read (cat) 1,573 0.8% Redirect-to-file, later read in full (“save then read”) 4,301 — common filter patterns guesses: Shape Count Likely context sed (bare, e.g. sed -n 'N,Mp' file) 15,123 line-range paging grep | head 9,497 search, then cap the results cd | head 5,998 cd dir && cmd \| head grep (bare) 4,523 direct search on a file uv | tail 3,864 tailing output of a uv run/uv pip command cd | tail 3,253 cd dir && cmd \| tail tail (bare) 2,764 log/output tailing ls | head 1,699 capped directory listing find | head 1,365 capped file search common slop guess counts: Metric Count Guess-then-give-up-and-read episodes (≥2 narrow attempts, then a full read) 541 Unresolved thrash episodes (≥3 narrow attempts, never followed by a full read) 2,701 Total narrow commands spent inside these episodes 5,006 Resolved episodes closed by the Read tool 447 (83%) Resolved episodes closed by a bare cat 94 (17%) multiple different cmd | filter guesses in a row here are full attempt-count distribution for episodes eventually resolving into a full read (example: running a command with a failed filter, then continuing to run cmd | filterA -> cmd | filterB over and over again until a “filter” matches instead of just saving output to a file and reading directly):
img,video{width:100%} claude-hooker The mystery of claude code over the past year and a half has been how does a world leading ‘AI productization lab’ continually generate such shitty amatuer globally distributed software and services over and over again? It’s almost like staffing a company entirely by 24 year olds seeing their net worth increase by $2 million per month for two years is a bad idea? Even the tech brain infosphere has noticed how claude code’s architecture is absolutely mornoic and amatuer and anthropic just continually ego-sociopath lies and nobody can stop them because hey, when you spend over $2 billion a month renting GPUs run by globally polluting gas turbines, they let you do it. don’t worry, bro, “it’s a javascript game engine in your console, ur not smaht enouf to understand our god-ai logic here!!!” as you may be able to tell, i’m deploying this post in annoyance/anger mode because I don’t understand why me, somebody who can’t afford a place to live and has 20 years of detailed technical experience, is giving free work to trillion dollar “ai” “startups” with good models but apparently abysmally amatuer software engineering practices. i guess it’s part of industry bubble syndrome. It’s not really worth thinking about what drives weird un-self-aware ego-sociopathic-autistic billionaires if you aren’t close enough to change anything in their minds though. The Bad i guess back to the point: given they have made claude so same-system dangerous with its cavalier attitude towards running destructive or system-resource-consuming actions you must install your own safeguards using claude code event hooks.
Kvestigate - Investigate Kubernetes Clusters Ever find yourself given access to a piece of garbage kubernetes cluster and you need need to audit it to figure out what the hell is going on? Say hello to kvestigate your kubernetes investigation and auditing toolkit with the ability to even run ad-hoc workloads beyond the reach of the kubernetes scheduler allocator substrate system by just doing shit on the hosts themselves like computers are supposed to do. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Language Files Lines Code Comments Blanks ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Python 141 34192 28551 1381 4260 TOML 3 72 57 8 7 YAML 23 1059 718 257 84 ───────────────────────────────────────────────────────────────────────────────── Markdown 18 3983 0 3083 900 |- BASH 13 500 288 141 71 |- Python 6 177 132 27 18 |- YAML 3 218 187 11 20 (Total) 4878 607 3262 1009 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Total 185 40201 29933 4908 5360 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Features kvestigate is a read-first Kubernetes toolkit for investigating clusters, running workloads directly on host machines (no Docker), and safely handing a constrained slice of your cluster resources to AI agents. Three surfaces on one foundation (the official kubernetes client + a privileged node-access DaemonSet): Investigate & plan — audit a cluster, stream logs, exec into pods, and bin-pack a workload onto available GPU/CPU. Read-only by default. Run things host-native — create isolated workspaces on a node’s host OS (tmpfs / disk / hybrid storage, GPUs bound by UUID), sync code in, and launch processes. No container build, no pod scheduling. Host constrained cyberspaces — expose an allow-listed, quota’d, audited slice of those workspace capabilities over HTTP/MCP, so agents and other programs can “upload and run things” without cluster-wide power. Bonus Feature Natively integrates with nvidia-smi to query all nodes for cluster-wide gpu usage and power metrics to see how much of the world you are wasting with your “ai” “workloads”
hyperdjango Following up from opsmas 2025 is spring cleaning 2026. Welcome! To kick off spring cleaning, the first project I’d like to release is: hyperdjango. What is hyperdjango you may ask? You could jump ahead into all docs at the hyperdjango hyperdocs hypersite and find out. Though, what you should be asking: what isn’t hyperdjango?
More in AI
Today's links On the sincerity of AI bosses: Fascism is always an incoherent bundle. Hey look at this: Delights to delectate. Object permanence: 9/11 v spam; PalmOS x WTC collapse; Wifi x WTC rubble; Berlusconi sex rings; Tesco bans writing down prices; AI psychosis and the warped mirror; Conspiratorialism's causal chain. Upcoming appearances: Budapest, Berkeley, Edmonton, Boston, South Bend, Hudson, Calgary, Winnipeg, Vancouver, Victoria, Ottawa, Montreal. Recent appearances: Where I've been. Latest books: You keep readin' em, I'll keep writin' 'em. Upcoming books: Like I said, I'll keep writin' 'em. Colophon: All the rest. On the sincerity of AI bosses (permalink) The word "fascist" comes from the Latin fasces, a bundle of sticks; the symbology here is that a single twig is weak and brittle, but bundled together, many twigs are strong. It's a sound political theory, because in politics, coalitions are everything: https://pluralistic.net/2025/01/06/how-the-sausage-gets-made/#governing-is-harder The problem with fascism isn't the idea of bundling together different groups: it's the incoherence of that bundle. The fascist coalition is a collection of people who want mutually incompatible things. When one part of the fascist coalition wins (say, if Nick Fuentes's neo-Nazis triumph), the other faction loses (Fuentes gets to murder Stephen Miller and turn his skin into a lampshade). The fascist coalition is a coalition of enemies who all hate each other and dream of exterminating one another, held in check by a strongman who uses flattery, favors and threats to keep a lid clamped tight on this pressure-cooker: https://pluralistic.net/2025/07/29/bondi-and-domination/#superjove In this regard, fascism is simply one end of the continuum of conservative movements, which are always about finding a way to "get turkeys to vote for Christmas." That's because, at root, conservativism is the belief that some minority (rich people, white people, bosses, men, etc) were born to rule and everyone else was born to be ruled over: https://pluralistic.net/2026/07/08/wilhoitian/#human-rights-v-property-rights By definition, "a minority that was born to rule" can't win an election, because they are a minority. Conservatives win electoral races by convincing people they intend to oppress, cheat and maim to vote for them through appeals to fear and hatred (racism, transphobia, sexism, anti-communism, etc): https://pluralistic.net/2022/03/09/turkeys-voting-for-christmas/#culture-wars Conservative political victories are always followed by economic misery for the conservative base, because the senior partners in the conservative coalition are the bosses who get richer by making workers poorer. Conservative rulers try to offset this with spectacular acts of cruelty against disfavored minorities, but this tactic only carries so far. Eventually, the electorate notices that despite terrorizing migrants and trans people, diesel is now $10/gallon and the guy responsible is now $1.4b richer than he was before the election: https://www.bbc.com/news/articles/cvgmv98ez3zo Workers and bosses aren't the only fracture line in the conservative coalition. Within conservativism, there are leaders who want mutually incompatible things and abhor one another: the white nationalists hate the Zionists; the misogynists hate the TERFs; the imperialists hate the isolationists: https://pluralistic.net/2024/07/14/fracture-lines/#disassembly-manual These fracture lines can be papered over while things are good, but they crack when things go wrong, and this is even more true of fascist movements than it is of other conservative coalitions. This is true of all fascists, so it's true of technofascists, too. The best-ever reference work on technofascism was just published: Naomi Klein and Astra Taylor's End-Times Fascism, which unpacks the apocalyptic ideology that dominates Silicon Valley, especially the AI cultists: https://naomiklein.org/end-times-fascism/ In a recent interview about the book with the QAA podcast, Astra Taylor explained how the contradictions of the technofascist movement are to be expected, because fascism is always an "incoherent bundle": https://soundcloud.com/qanonanonymous/end-times-fascism-feat-naomi Understanding technofascism's inherent incoherence is vital to making sense of the chaos roiling the AI cult at this moment, wherein you have AI people insisting that there must be a moratorium on AI development lest the word-guessing program awaken and devour the human race. This week on the Better Offline podcast, Ed Zitron discussed the outlandish, science-fiction inspired cult beliefs that dominate AI boardrooms with Adam Becker and Cal Newport: https://www.youtube.com/watch?v=0oVSnaINJ30 Becker is well-placed to discuss this. Like the hosts of the QAA podcast, he started paying close attention to the bizarre beliefs of conspiratorialists long before the rest of us realized that no matter how preposterous their certainty about the imminent machine intelligence Singularity was, these beliefs are sincerely held by some very wealthy and driven people. Becker's 2025 book More Everything Forever is a tremendous field guide to these delusions and their profound philosophical and technical deficits: https://pluralistic.net/2025/04/22/vinges-bastards/#cyberpunk-is-a-warning-not-a-suggestion In the interview, Newport dismisses the theory that the warnings about imminent AI apocalypse are self-serving criti-hype intended to serve as both marketing pitch and regulatory capture gambit, through which the hyperscalers get the government to step in to interrupt the beggar-thy-neighbor doom-loop: https://pluralistic.net/2026/09/16/beggar-thy-neighbor/#red-queens-race Rather, Newport says that these people sincerely believe that they are about to immanentize the eschaton and are pants-wettingly terrified about the AI god they will conjure forth any day now. He makes a good case for this, pointing to the long history of words and deeds on the part of various AI bosses that suggest that they are true believers who are genuinely high on their own supply. I don't doubt that there are sincere believers in the AI technofascist coalition, but that does not preclude the possibility that they share their boardrooms and executive rows with cynics for whom this is all a shuck, a scare-story to convince the rubes that their modestly useful utility software is really a nascent "superintelligence" and thus capable of replacing all their workers, which means they should fire all those workers and start sending their salaries to AI companies. This is an example of one of those "incoherent fascist bundles." Just as Mike Pence (a misogynist Christofascist) was happy to share the White House with Trump (a godless pedophile rapist), AI companies can and do thrive by filling their executive ranks with Singularity-crazed maniacs and sharp operators who are happy to spread this superstitious nonsense if it helps them pump up their stock swindle. Each group thinks they're using the other one, and they are…up to a point. When it comes to the current AI nonsense, that point came when Nvidia's best customers started to demand that everyone stop buying Nvidia's products, whereupon Nvidia's CEO suddenly remembered that his chips weren't being used to make god, but rather, to power regular-degular "cloud software": https://cxotoday.com/governance/nvidias-jensen-huang-crosses-swords-with-ai-labs-over-regulation/ When it comes to technofascists (and all fascists) this kind of division isn't an exception, it's the rule. The billionaires behind AI are split between solipsists who don't believe other people are any more real than bots; and cynics who think that bosses will be easy marks for a sales pitch that sees them replacing mouthy workers with pliable chatbots: https://pluralistic.net/2026/08/03/andor/#either To be a senior member of the fascist coalition, you must be capable of both sincere belief while not openly dismissing your fellow senior members' contradictory sincere beliefs. Behind closed doors, they may make fun of each other (or fantasize about murdering one another), and they may periodically erupt into plots to oust one another from the coalition. But every one of them must be able to go along to get along… Most of the time. Until they don't. Hey look at this (permalink) The High Crime of “LMAO”: How Cops Are Treating Mass Surveillance As a Joke https://www.eff.org/deeplinks/2026/09/high-crime-lmao-how-cops-are-treating-mass-surveillance-joke Rethinking space opera https://www.antipope.org/charlie/blog-static/2026/09/rethinking-space-opera.html EU wants Canada to become ‘associate member,’ von der Leyen says https://www.politico.eu/article/eu-wants-canada-to-become-associate-member-von-der-leyen-says/ The Trump Administration Creates a Monopolization Machine https://prospect.org/2026/09/16/trump-administration-creates-monopolization-machine-small-business/ a bad tool always blames the workman https://backofmind.substack.com/p/a-bad-tool-always-blames-the-workman Object permanence (permalink) #25yrsago 9/11 v spam https://memex.craphound.com/2001/09/17/through-most-of-last-week/ #25yrsago PalmOS picture of the WTC collapse https://web.archive.org/web/20010920145653/https://ne.nikkeibp.co.jp/english/2001/09/0914pda_watch.html #25yrsago Wifi emanating from the WTC rubble https://web.archive.org/web/20010916231834/http://dailynews.yahoo.com/h/nm/20010916/tc/attack_wert_dc_2.html #15yrsago Silvio Berlusconi prostitution-ring wiretaps: sex with eight women in one night, “I’m only prime minister in my spare time” https://www.theguardian.com/world/2011/sep/18/silvio-berlusconi-wiretaps-sex-parties #15yrsago Tesco threatens journalist with arrest for writing down prices https://www.theguardian.com/money/blog/2011/sep/16/tesco-shopping-supermarket-prices-check-writing #1yrago AI psychosis and the warped mirror https://pluralistic.net/2025/09/17/automating-gang-stalking-delusion/#paranoid-androids #1yrago Conspiratorialism's causal chain https://pluralistic.net/2025/09/17/cause-and-effect/#things-have-causes Upcoming appearances (permalink) Budapest: Brain Bar, Sep 17 https://brainbar.com/munkatars/cory-doctorow Berkeley: Celebrating 25 Years at the Digital Frontier (Samuelson Clinic), Sep 24 https://www.law.berkeley.edu/experiential/clinics/samuelson-law-technology-public-policy-clinic/samuelson-25th-anniversary-celebration/ Edmonton: Elbows Up (Edmonton Public Library), Sep 28 https://www.epl.ca/blogs/post/elbows-up-with-cory-doctorow/ Boston: The Post-American Internet: Possibilities for a new internet created by an American Hermit Kingdom (MIT Media Lab), Sep 30 https://www.media.mit.edu/events/the-post-american-internet-possibilities-for-a-new-internet-created-by-an-american-hermit-kingdom/ Boston: The Paradox of Enshittification and Reverse Centaurs (Harvard Berkman Klein), Sep 30 https://cyber.harvard.edu/events/running-harder-falling-faster-paradox-enshittification-and-reverse-centaurs South Bend: An Evening With Cory Doctorow (Notre Dame), Oct 6 https://franco.nd.edu/events/2026/10/06/an-evening-with-cory-doctorow/ Hudson, OH: Hudson Library, Oct 7 https://engagedpatrons.org/EventsExtended.cfm?SiteID=3850&EventID=596952&PK= Calgary: Wordfest, Oct 8 https://wordfest.com/2026/show/wordfest-presents-cory-doctorow-2026/ Winnipeg: McNally Robinson, Oct 9 https://www.mcnallyrobinson.com/event-18991/An-Evening-with-Cory-Doctorow Vancouver: Read, Resist, Repair, Rejoice (Vancouver Writers Festival), Oct 19 https://writersfest.bc.ca/festival-event-2026/01 Victoria: Munro's Books, Oct 20 https://www.munrobooks.com/events/6113620261020 Vancouver: Life After AI (Vancouver Writers Festival), Oct 22 https://writersfest.bc.ca/festival-event-2026/46 Ottawa: Life After AI (Ottawa Writers Festival), Oct 24 https://writersfestival.org/event/life-after-ai Vancouver: BC Policy Solutions Gala, Nov 12 https://bcpolicy.ca/gala/ Montreal: World Science Fiction Convention, Sep 2-6 https://montreal2027.ca/en Recent appearances (permalink) Could Tech Bosses Destroy Life As We Know It? (Politics JOE) https://www.youtube.com/watch?v=PL4VktU0SgY Are 'AI Apocalypse' Warnings Just Marketing? (What's Left) https://www.youtube.com/watch?v=IXd9HwIE5bo The Real AI Threat Isn’t What You’ve Been Told (The Tea with Myriam François) https://www.youtube.com/watch?v=Vc8It00fRsA Fascists may come after the AI bubble bursts (You&AI) https://www.youtube.com/watch?v=J2WN64aQeYQ What Would a Normal Person Do (Trashfuture) https://www.patreon.com/trashfuture/posts/what-would-do-169247456 Latest books (permalink) "The Reverse-Centaur's Guide to AI," a short book about being a better AI critic, Farrar, Straus and Giroux, June 2026 https://us.macmillan.com/books/9780374621568/thereversecentaursguidetolifeafterai/ "Canny Valley": A limited edition collection of the collages I create for Pluralistic, self-published, September 2025 https://pluralistic.net/2025/09/04/illustrious/#chairman-bruce "Enshittification: Why Everything Suddenly Got Worse and What to Do About It," Farrar, Straus, Giroux, October 7 2025 https://us.macmillan.com/books/9780374619329/enshittification/ "Picks and Shovels": a sequel to "Red Team Blues," about the heroic era of the PC, Tor Books (US), Head of Zeus (UK), February 2025 (https://us.macmillan.com/books/9781250865908/picksandshovels). "The Bezzle": a sequel to "Red Team Blues," about prison-tech and other grifts, Tor Books (US), Head of Zeus (UK), February 2024 (thebezzle.org). "The Lost Cause:" a solarpunk novel of hope in the climate emergency, Tor Books (US), Head of Zeus (UK), November 2023 (http://lost-cause.org). "The Internet Con": A nonfiction book about interoperability and Big Tech (Verso) September 2023 (http://seizethemeansofcomputation.org). Signed copies at Book Soup (https://www.booksoup.com/book/9781804291245). "Red Team Blues": "A grabby, compulsive thriller that will leave you knowing more about how the world works than you did before." Tor Books http://redteamblues.com. "Chokepoint Capitalism: How to Beat Big Tech, Tame Big Content, and Get Artists Paid, with Rebecca Giblin", on how to unrig the markets for creative labor, Beacon Press/Scribe 2022 https://chokepointcapitalism.com Upcoming books (permalink) "The Post-American Internet," a geopolitical sequel of sorts to Enshittification, Farrar, Straus and Giroux, 2027 "Unauthorized Bread": a middle-grades graphic novel adapted from my novella about refugees, toasters and DRM, FirstSecond, April 20, 2027 "Enshittification, Why Everything Suddenly Got Worse and What to Do About It" (the graphic novel), Firstsecond, 2027 "The Memex Method," Farrar, Straus, Giroux, 2027 Colophon (permalink) Today's top sources: Currently writing: “Once Is Enemy Action,” a science fiction novel about the origins of modern technofascism. Today's words: 501 (15980 total). "The Post-American Internet," a sequel to "Enshittification," about the better world the rest of us get to have now that Trump has torched America. Fourth draft completed. Submitted to editor. A Little Brother short story about DIY insulin PLANNING This work – excluding any serialized fiction – is licensed under a Creative Commons Attribution 4.0 license. That means you can use it any way you like, including commercially, provided that you attribute it to me, Cory Doctorow, and include a link to pluralistic.net. https://creativecommons.org/licenses/by/4.0/ Quotations and images are not included in this license; they are included either under a limitation or exception to copyright, or on the basis of a separate license. Please exercise caution. How to get Pluralistic: Blog (no ads, tracking, or data-collection): Pluralistic.net Newsletter (no ads, tracking, or data-collection): https://pluralistic.net/plura-list Mastodon (no ads, tracking, or data-collection): https://mamot.fr/@pluralistic Bluesky (no ads, possible tracking and data-collection): https://bsky.app/profile/doctorow.pluralistic.net Medium (no ads, paywalled): https://doctorow.medium.com/ Tumblr (mass-scale, unrestricted, third-party surveillance and advertising): https://mostlysignssomeportents.tumblr.com/tagged/pluralistic "When life gives you SARS, you make sarsaparilla" -Joey "Accordion Guy" DeVilla READ CAREFULLY: By reading this, you agree, on behalf of your employer, to release me from all obligations and waivers arising from any and all NON-NEGOTIATED agreements, licenses, terms-of-service, shrinkwrap, clickwrap, browsewrap, confidentiality, non-disclosure, non-compete and acceptable use policies ("BOGUS AGREEMENTS") that I have entered into with your employer, its partners, licensors, agents and assigns, in perpetuity, without prejudice to my ongoing rights and privileges. You further represent that you have the authority to release me from any BOGUS AGREEMENTS on behalf of your employer. ISSN: 3066-764X
There's a lot of polarising discourse right now about the threat AI poses to humanity. Some think it's a farce and others think we face extinction. Here are my thoughts.
A middle ground between the cybersecurity and AI safety communities
An overview of the current state of the engineering market and the AI skills that are in demand
