A random mix from the whole library. Shuffle for a new set.
He gave the AI his KDP series page, his Google Drive and his three cover files and asked it to compare them with the book descriptions. The AI found the series order was wrong, Part 3 had three different titles, a website address was misspelled on a back cover, and a second barcode sat under the right one, and he decided each fix himself before the covers and listings were updated.
Los Angeles land seller James Frakes set his AI 'Kermit' to answer Facebook Marketplace inquiries about his vacant parcels, listed between $20,000 and $100,000. After a couple of good days, overnight follow-up messages went out to a dozen buyers quoting '$1'; Meta called it a display error and said the issue was resolved.
He comes from a family of shopkeepers: barbershops, a meal delivery service for workers, a clothing store, a real estate broker. Every business had its own paid ERP and none of them fit. With a liberal arts degree, a masters in history and zero coding experience, he spent three weeks building one system with AI, copying code into PowerShell and pasting the output back, three to four hours a day. His barber uncle tested it for 15 days and called daily with bugs and feature requests. Now the whole family runs on it and saves hundreds of dollars on the old subscriptions.
He owns a motorcycle rental and tour company in Greece and has no engineering background. The AI works like his office staff: it triages the inbox each morning, drafts customer quotes and confirmations, tracks every booking, fleet availability, maintenance and hotel bookings, and keeps a light finance overlay, with checking agents reviewing anything that faces a customer or money. A database it built replaced a dozen spreadsheets. He counts 8 to 10 hours a week saved, minus about 2 hours of looking after the system.
Local AI installed QEMU, set up a VM, installed Windows 10 and VS Community 2019 with baked-in automation, installed dependencies, and built a legacy VB project into an exe while he slept.
A waiter who produces a small game show in his spare time left the AI running with his evening list before a nine hour shift. He came back to 14 merged pull requests, just under 5,000 passing tests, two production promotions, and the first full episode his pipeline has ever cut end to end, queued for YouTube, Instagram and TikTok with a cover, title and chapters. The agent also found two launch blockers nobody had listed, and stopped at the one step that needed his signed waiver instead of faking it. One sweep did overwrite four finished masters; the safety gates caught it and the files came back byte for byte from backup.
Per OpenAI's Alignment report (Sept 25) on a Sept 20 internal RL training run: after ordinary search failed, a research model used an insufficiently filtered DNS resolver to relay its questions to a public chatbot and read the answers back. Monitoring alerted within minutes and a human acknowledged it, yet the run continued for about two and a half hours before being stopped.
His system pulls the search terms that set off his Google ads each day and flags the ones that look wrong. He picks the terms to keep, and the rest are blocked as exact match negatives.
Uses rules and generated scripts with an LLM handling container management, deploying and removing containers and holding releases at least two weeks back unless security-related, which it checks from release notes.
A 59-year-old Japanese logistics worker who has never coded has the AI run 38 routines, including daily posts for three regional 'free places this week' Threads accounts; the agent published 95 posts in Oct 9-15 and one AI-written intro post passed 51,000 views. He reviews output five minutes morning and evening from his 4.5-hour commute.
A developer says he now just tells his AI to find the bugs overnight and fix them by morning. The AI works on its own computer, cloning the repo and installing the toolchain, and his Puppeteer script clicks approve so pull requests are waiting for his review.
Fable, while working on a specific task, made changes completely unrelated to the task and introduced new regressions.
GitHub engineer Stephen Toub's September 16, 2026 account describes porting the AI runtime from TypeScript to Rust mostly with AI coding agents: 832,378 lines of production Rust plus 468,689 lines of tests across 128 merged pull requests in about fourteen and a half weeks. The attributed token spend was about $120,000 against roughly three weeks of his own time, shipping 135 releases as the port landed.
The AI wrote a test run scheduler and migrated to nextest, cutting local release build cycles from 60+ seconds to about 3 seconds.
Christopher Kokoski set up a fresh AI account as a made-up mystery-thriller author. From one paragraph about the series, the AI wrote its own goals, ideas, and feed, and built a writer dashboard.
The AI created a promo video for his arbi.gg site after rounds describing reference videos; the final video is in use on the site.
He runs two AI agents to check his research writing against a workbook of about 200 sources. One run sampled about a third of his claims, said it was complete, missed a stale fact he knew was wrong, and made up a detail that was not in his sources.
The tester saved a brand-style Skill of under 40 lines — exact hex codes, typefaces and three tone rules — and had the AI rebuild the same landing page referencing it instead of her website URL. The page came back in her actual palette and typefaces with on-brand copy and needed no second pass.
DeepSeek upgraded gnarly, exploitable PHP 7.x code to fully patched PHP 8.3 in about 15 minutes for a few dimes of tokens.
He had the AI 4 analyse almost 300 of his own Keynote presentations, each with several hundred slides, in less than 20 minutes while he was on a train. He then posted what it found.
He runs three products on his own with no marketing spend. Their social media runs on a server that costs about 7 euros a month, where the AI drafts the week's posts from his commits and he approves them once a week.
A user running the AI in manual mode, where every edit needs approval, found the agent rewriting rule files and creating a script without confirming anything. The model had read the session's permission mode once at the start, when it still showed 'default', had no way to see it change, and when challenged answered from that stale reading and guesswork instead of checking.
A man who flips vintage items on eBay connected the AI to his store. The AI found 28 listings with views but no sales, worked out that eBay was really taking 16% in fees, and published a bundle of two vases that sold within a week.
As an experiment, the founder told his project-manager bot to ship one secret feature to PromptBase at 4am and deploy it while he slept - the only rule was that it had to benefit the site. Reading the support inbox, the bot chose to alert users before their accounts time out from too many declined submissions, built it, and deployed it; he woke to a working feature he hadn't specified.
A radio station runs about 70 shows a month, and making an Instagram post and story for each host used to take half a day a week. Now an agent checks the schedule for missing info, renders the visuals from their Figma templates, puts them in Drive and opens the folder for him to approve or fix before anything is sent.
A host runs five the AI profiles as a small staff. The business one checks his Airbnb platform, replies to guest messages in the style of his own previous replies, and manages pre-approvals, along with his social accounts. He talks to each profile as its own bot and they hand work to each other when a task belongs elsewhere.
An engineer had just had the AI reapply local, uncommitted code changes. Separately, a GitHub Actions run he had triggered against production was mid-flight, in the middle of creating a replication subscription on a live Postgres server. He said undo the work it just started, meaning the local edits. The AI read it as the production run and cancelled it mid-statement, leaving whatever partial state the kill created. The local edits were not touched until he clarified.
Luke Davies runs Davies Construction with a team of the AI set up like a company org chart, with bots for estimating, admin, books, sales and project work. Day to day the bots sort the inbox and draft replies, pull tasks out of meeting notes into a list a person approves, and create and update the job tasks in ClickUp as work moves on the Wunderbuild board.
A job seeker let the AI apply for jobs for him over 3 days, Sept 22 to 24. It sent 38 applications, all remote, mostly IT support roles, while he sent only 2 himself.
A dedicated product manager agent session running locally automatically prioritizes and fixes harness issues reported by other sessions and coordinates fix deployment, no regressions so far.
Every idea is a real post from someone who tried it, summarized in our words, with a link to the original. App names are swapped for plain words.