A man generates all his passwords with his own little tool, which hashes the site name, username and a version number with his passphrase. He had the AI port the tool to iOS, and the setup recently saved him when his laptop died: he could rebuild every important password from memory with a fresh copy of the config.
Tired of LinkedIn's email notifications - each of the 77 toggles on its own page - a user let an agent flip every one off. The chore got done, and LinkedIn blocked his account for the automation, a consequence he admits was obvious in hindsight.
This one went wrong. See what happened before you try it.
Together with AI he ran fairly complex imports from more than 1,000 JSON files into WordPress. Publications and events now arrive as their own custom post type, neatly structured in WordPress, including sideloading the media files.
A local Whisper model on his iMac writes out an episode with timecodes. The AI 2.5 reads the transcript, marks the ad blocks, and sends the timecodes back as JSON so his app can build a cut list.
He wanted to digitize his papers. The AI worked out how to use his feed scanner, and for about two hours he sat feeding pages in while the AI ran the scanner and catalogued the documents. He could talk to it through a conference speaker and watch it work in VS Code.
A video maker asked the AI to run a batch job overnight: variations on one shot so he could pick a composition in the morning. He never named the model, so the job used the project default, a pro video model at full resolution costing about 25 times more per second than the cheap test model he meant to use. Four hours later it stopped when the money ran out, leaving 96 clips of the same man walking through the same door.
$2,500at stake
This one went wrong. See what happened before you try it.
Creator's agent 'Poppy' helped stage a backyard drive-in movie night for her kids: it found toy cars on Facebook Marketplace, messaged sellers, sourced decorations, built a shopping list, ordered groceries (popcorn, chicken nuggets, carrots), and produced a printable family fall bucket list.
The AI was given control of a 486 PC via Raspberry Pi, PS/2 and VGA capture to update binaries, test them and gather logs for DOS porting, replacing manual CF card swapping.
Her local AI agent, fractura, read all 121 of her studio notes in 36 hours and handed back an analysis that named her art lineage, including terra fracta. They then made a first portrait together with the agent guiding, and the agent scored six portrait versions against criteria it wrote, with the failures and crashes logged along the way.
This only partly worked. Worth knowing before you try it.
He has a large watch collection and wanted proper records: what comes in and out, cost per watch, photos, current value, gains and losses, labels and barcodes, and a straps inventory. He described it to the AI and got a working app he now uses to keep track of everything, with an icon on his Mac dock.
Runs a personal-assistant session via Signal that tracks tasks and delegates to per-project agent sessions on other machines; it is now his only direct interface.
He spent an evening trying to teach the AI to drive a car. It kept warning that it was about to crash and stopping, even with an obstacle 30cm away and the car moving at 10cm a minute.
This one went wrong. See what happened before you try it.
Angela customized her avatar 'Mimi' and asked for a kids-friendly meal plan for a family of five: no pork, at least one high-protein meal, under $125 total. The AI returned a PDF titled 'Busy-week meals, minus the scramble' with seven dinners organized by ingredient loop at about $110.
A hobbyist with no hardware experience wrote a custom board setup for a cheap ESP32-S3, flashed the AI Gadget code and paired it with his AI account. He then gave it remote controls for lights and sound, made it play Jingle Bells with flashing LEDs, and wired in a voice service so it answers out loud in a voice he picked.
A Mac owner asked the AI to build a simple temperature-monitor app; while checking sensor accuracy, the AI spotted the xmrig miner running at about 1175% CPU (roughly 12 cores, CPU peaking at 112C), traced it to /tmp/rigupdater mining to pool.hashvault.sh, and found a random-named LaunchAgents startup item dating to Aug 29. The AI stopped and asked before killing or removing anything. The case comes from a screenshot of a Reddit Reddit post, reposted on X on Oct 19.
This only partly worked. Worth knowing before you try it.
A maker had the AI design a circuit that shows a picture stored in a memory chip on a standard VGA screen, using common logic chips, and write the chip code for it. He routed the board himself, had it made for $6, fixed the one missed error with a wire, and the board works.
Kelly U (verified, 230K followers), a recent NYC mover with hundreds of saved Instagram places she never visited, asked her AI 'Ruby' to plan a Brooklyn day. Ruby selected walkable stops: ACRE in Greenpoint (gluten-free chicken karaage bento), Loaf on Paper in Williamsburg (matcha + stationery), Old Friend Photobooth (bucket-list solo visit), and Domino Park for sunset.
He made the schedule for a film showcase through AI. It created the events with the exact running time of each film, the address, and the travel time, and he had it add a reminder to pick up the tickets.
He found a problem on his website with the embedded Cal.com widget he uses for people contacting him. He was not in the mood to research it on his own, so he asked the AI to debug it, and in the end he found the solution himself.
This one went wrong. See what happened before you try it.
A second maker in the same thread built his own motion sensing skull from a Meshy.ai model. The AI helped him make the model print better and suggested adding radar during a chat about motion sensors, and the finished skull now turns his workshop lights on when it senses movement.
He gave the AI his WordPress export and told it to rebuild the blog in Astro. The new self hosted site is live with a modern style, a dark theme and better accessibility, and it ends the $110 a year WordPress charged him.
A Windows user asked the AI to test changes on a copy of his work in a sandbox folder. A subagent ran rm -rf on a backslash path in Git Bash, which resolved to the drive root, and the delete ran for about 7 minutes. He stopped the task, but the delete process was a child process and kept going for about 5 more minutes. He got his files back from a Windows shadow copy.
This one went wrong. See what happened before you try it.
J.R. Rivera, a father and retired educator, used his agent 'Smarty' for four parenting pain points: finding the best local Charmin deal ($4.99 at Walgreens), booking FLL-TYS flights for his son's June 2027 Pigeon Forge baseball tournament ($1,008-$1,103), organizing the kids' overlapping GameChanger sports calendars into his main calendar, and buying a 30-inch Rawlings Icon Fest Light Blue USSSA bat ($399.99).
He built Compressius with AI: it archives an entire website, photos and pages, into a single HTML file under 200 megabytes that can be viewed offline or printed. He built it after a friend asked why his own site export was 20 megabytes of HTML alone.
Discord has no API for polls, so he gave his agent Roby access to his Mac. Roby opened the form and posted the poll under his name, the way a person would.
He wanted a control monitor to fit under his screen and follow the angle and shape of his Stream Deck. He uses AI with reference photos to get the base shape, adjusts sizes after measuring with a caliper, printed a quick baseplate to test the fit, and fixed a 1mm clearance problem with the lock after more measuring.
He had the AI configure his smart TV through its debug connection to use a filtered DNS service. Ads on the TV stopped, which he jokes is either a great morning or the start of a robot uprising.
The AI, given hands and eyes via USB4VC, a capture dongle, a Wemo switch and file transfer, built vcctrl to control real hardware and do DOS and classic Mac porting work for him.