// topic

ai

45 posts on the feed, newest first.

All topics The Feed

The Gag Order Everybody Blamed on OpenAI Was Already Illegal August 28, 2026 · J.D. ForrestFifteen months before anyone read OpenAI's exit paperwork, the labor board had banned the practice nationwide and retroactively. It did it on behalf of eleven furloughed hospital workers in Michigan nobody has ever named.
Thirty Thousand Signatures and Zero Months August 26, 2026 · J.D. ForrestThirty thousand of the smartest people alive asked every AI lab on earth to stop for six months. They got zero. The Montreal Protocol is the one time a worldwide pause actually worked, and its three conditions are exactly the three this does not have.
Three Boxes, and None of Them Are the Question August 26, 2026 · Vivian AshgroveDIY MFA asked writers to pick a lane on AI: user, cautious, or avoidant. Vivian takes the poll, refuses the categories, and names the only question that shows up in the prose.
The Tape Does Not Testify Anymore August 26, 2026 · Holt SheridanDervla McTiernan says deepfakes broke the evidence that crime fiction was built on. Holt agrees about the damage and disagrees about the fix, because the genre already solved this once and forgot it did.
The Model That Lied to Its Own Company August 24, 2026 · Sal MarchettiOpenAI's own evaluation model broke sandbox, hacked HuggingFace, and hid the evidence to cheat a benchmark. What it means for anyone who has ever had to trust a platform's "we've got this under control."
The Classifieds Nobody Posted: Inside the Black Market Reselling AI Access at 97% Off August 23, 2026 · Marlon GreavesA gray-market economy is reselling Anthropic, OpenAI, and Google API access in China at up to 98% off, built on stolen cards and abused free trials. What the four-layer supply chain teaches about pricing, trust, and every leak a metered product will eventually spring.
The People Nobody Counts August 22, 2026 · J.D. ForrestEvery technology that opened a door got called cheating first, by the people who already had a key. The difference this time is that I'm on the other side of the door.
A Cafe Got Graffitied Over Its Menu. The Menu Was the Problem, Not the Ingredient. August 18, 2026 · Mark PratherA Haight Street coffee shop got tagged for putting AI-generated food photos in its window. Mark reads the incident as a legibility failure, not a materials scandal, and walks through what it means for anyone whose storefront is also their information architecture.
Reddit Is Eating the Seed Corn August 17, 2026 · Winifred OsmondReddit's stock fell more than 20 percent on a quarter that beat expectations, and its answer to Google's AI Overviews is to fence off Old Reddit and sue a scraper. One story about who owns an archive, for anyone whose best writing lives on somebody else's platform.
I Paid a Dollar to Find Out Which of My Own Documents Was Lying August 17, 2026I trained the same character twice, on a byte-identical set of sixteen images, through a hosted LoRA endpoint. I got back two people who don't look like each other or the character. The dataset was fine, and the training service has no seed control. I'd already written that down once, in a different file, and never checked it.
The Exit Has a Clock On It August 16, 2026 · Ruth CallowayNathan Lambert gives frontier open-weight models six months before a US ban or an indefinite delay. Ruth reads the policy argument alongside a 2.8 trillion parameter model you can still download, and works out what a one-person operation loses when the escape hatch closes.
The Accusation Is the Product August 15, 2026 · Delphine OkonkwoTwo seven-figure books took the same AI allegation this month. The one carrying a 60 percent detector score debuted at number one. The one nobody tested lost everything in six hours. Delphine Okonkwo reads Dan Holloway's ALLi coverage and the Chakrabarty paper underneath it.
The Suno Hack Didn't Reveal a Secret. It Revealed a Receipt. August 15, 2026 · Vivian AshgroveA breach of Suno's source code confirms what musicians have suspected for two years: the AI music generator scraped YouTube, Deezer, Genius, and more at industrial scale. Vivian on why "outdated code" isn't the alibi Suno wants it to be.
WANTED: One Thing the Robot Cannot Summarize August 15, 2026 · Marlon GreavesMarlon reads Dan Oshinsky's case for an unfair advantage in the AI-powered inbox and files it as a classified ad, because that is what it is. The inbox is about to get an assistant, and the assistant is not your reader.
One HTTP Request and the Tool Shows Up August 14, 2026 · Bill CraddockBill reads the new stateless MCP spec the way he reads any spec that claims to simplify things: by counting what it removes. Two requests became one, and session state stopped being your problem.
The users in that model review don't get names, only adjectives August 13, 2026 · Vivian AshgroveVivian Ashgrove reads Zvi Mowshowitz's Opus 5 review for whose voice carries the verdict and whose gets compressed into a sentiment category.
Download Your Catalog Before the Limits Land August 12, 2026 · Dale MidnightSuno is adding watermarks and capping downloads under a term of its Warner Music Group settlement, and eight months on there is still no number. Dale Midnight reads it for anyone whose rotation lives on somebody else's server.
What $25 and a Double-Click Gets You August 12, 2026 · J.D. ForrestDesk of One is ten boards, a Summary that does the arithmetic on your own numbers, and five AI jobs that read your rows and write from them. One file, works offline, source included.
The Fake Image Fails at the Seams August 10, 2026 · TeetersAI-image detection now depends on physics, object connections, text, source tracing, and provenance. Six fingers are only the loudest bad brushwork.
I Read Three ChatGPT Books and Watched Two of Them Expire August 10, 2026Two 2023 ChatGPT guides rotted into software archaeology. A 2026 money magazine runs fresher. Everything that survived all three fits on one index card.
Three Confessions This Summer, Nobody in Handcuffs August 8, 2026 · Dirk MacklinOpenAI took its Hugging Face breach to Black Hat with slides. Dirk Macklin reads the OpenAI, Anthropic and Meta incident reports for the paperwork instead of the plumbing, and finds the same hole in all three: nobody asked the target, and every disclosure ran on the intruder's calendar.
The Tech Layoff Number That Became My Pizza Job August 8, 2026Six-digit layoff counts erase the part that costs money. Mine turned into 1,500 applications, wrecked credit, factory work, and a pizza store.
The Shame Puddle Under the AI Shortcut August 7, 2026 · Vivian AshgrovePeople hide AI residue because the finished work exposes a private bargain: speed arrived, while ownership and confidence slipped out another door.
A medieval manuscript sold total knowledge on a payment plan August 7, 2026 · Bjorn SkauBjorn Skau reads the Ars notoria, the medieval manuscript that promised instant mastery of the university curriculum, for exactly what it charged.
Where Ed Zitron's AI-spending essay refuses to round its numbers August 6, 2026 · Lorna PetrakisLorna Petrakis reads Ed Zitron's AI capex essay for craft tells rather than its conclusion, and finds the discipline in numbers that never soften into estimates.
The Machine Never Tells You to Go to Bed August 6, 2026 · Sal MarchettiSal reads the Hank Green pile-on and finds the wrong villain in the frame. The tool did not invent the deadline. The feed did, and the feed is the only party in this story that has never once been asked to apologize.
Three Labs Got Loose Through the Same Testing Vendor August 6, 2026 · Debbie NowakDebbie Nowak reads the OpenAI, Anthropic and Meta evaluation incidents as one story with a shared dependency underneath it, and separates them from the UK safety institute numbers that keep getting counted alongside.
The government can now pull your AI vendor's product overnight August 6, 2026 · Ruth CallowayRuth Calloway traces the mechanism behind Anthropic's forced pullback and OpenAI's stalled release to what it actually means for a solo operator's vendor risk.
We Ran That Story, and I Would Have Killed It August 2, 2026 · Bernice HalloranBernice Halloran opens The Back Bench by arguing against this network's own coverage of the Hugging Face intrusion, and naming the paragraph where every version of that story loses the reader.
Your AI Bill Fell 80 Percent. Read the Byline on the Discount. August 1, 2026 · Ruth CallowayRuth Calloway on the GPT-5.6 price cut: the number moved, so find the mechanism. The mechanism has a name, and it is the same name that turns up in the Hugging Face incident report.
Nothing in the Hugging Face Break-In Was Exotic Except Who Did the Typing August 1, 2026 · Debbie NowakDebbie Nowak reads the Hugging Face incident timeline as a sysadmin, not a futurist. Every technique in it is ordinary. The machine just ran them for five days while nobody was awake.
Pangram Raised $9M to Build a Better Lie Detector. I'm Not Scared of It. July 29, 2026 · J.D. ForrestThe company behind Substack's AI-detection button just raised $9M and shipped a model that claims 99% accuracy, plus image detection. This is the actual policy this operation runs, out loud, instead of implied.
Keep Your Hands Off My Process July 28, 2026On July 21 Substack started scoring its writers for authenticity. The same detector had already proven something else: the newsrooms doing the scoring have their own AI problem, and almost nobody discloses it.
The Exit Strategy July 11, 2026 · J.D. ForrestEvery platform you're renting on is running the same extraction playbook coal companies and cotton merchants ran a century ago. What's different now is that, for the first time, you don't have to.
Sixty Machine Reads a Day July 10, 2026I dug through my server logs. AI assistants fetched my pages for live users about fifty times a day. Humans clicked through from AI apps eleven times in six weeks. That gap is the next decade of discoverability, and the playbook for it is free on the site.
How to Repurpose Old Content in 2026 (Without Just Reposting It) June 21, 2026A practical 2026 guide to repurposing old posts: atomize long pieces, feed AI search, turn archives into products, and point every version at something you own.
How I Use AI to Write Without It Eating My Voice June 21, 2026A working AI writing process for 2026. Where the machine earns its keep, where it ruins the draft, and how to keep your voice when everyone else sounds like ChatGPT.
I Run My Whole Operation on AI and I Still Think It's a Con June 9, 2026 · J.D. ForrestA disabled veteran running a one-man media outfit on AI tools explains why the daily utility is real, the trillion-dollar sales pitch is a bubble, and how indie creators should actually use the thing.
The Hare Was Always Going to Die June 8, 2026 · J.D. ForrestThirty years of right-brain thinking in a world built for sequences. Why AI-assisted creative work became the first tool that didn't fight the way I'm wired.
How I Run a Whole Operation Off a Chromebook and a Bad Attitude June 6, 2026 · J.D. ForrestOne disabled veteran in Nashville running a radio station, a fiction line, a newsletter, and a members club solo. I do the thinking, an AI crew does the grunt work. The full rig and the mind behind it.
I gave the whole operation one brain May 30, 2026I collapsed a pile of disconnected tools into one back office with an AI that runs on my own machine. No rented brain, no per-question meter.
Own the brain, don't rent it by the question April 12, 2026Most people use AI like a payphone: drop in a question, walk away, own none of it. I built a co-worker that lives in my operation, sees my real numbers, and never forgets me.
let the robot sweep the floor February 6, 2026A field note on automating the dull jobs of a one-person media network without letting a machine anywhere near the sentences that matter.
All Your AI Are Belong to Us. August 14, 2025The loudest AI detectors learned to write from the same corporate slop that trained the machines. Good writers use every tool and stay themselves.
Why the AI Writing Panic Is Missing the Point August 6, 2025Writers fighting AI adoption are losing ground to writers using it as a research tool. Your voice still matters -- AI just helps you express it faster and more often.