An HVAC company owner who runs several small businesses counted 27 active AI agents and 36 routines, 26 of them scheduled, working across his companies: pulling statements, filing invoices, updating sales and cost sheets, flagging overdue maintenance and tracking renewal dates. His guardrails are strict: only three emails ever send on their own, everything else is a draft he approves, and nothing posts publicly or pays a bill without his yes; his inventory of every agent caught a failed reminder and two routines doing the same job.
Dylan Heagy photographed a fridge whiteboard and asked the AI to order the grocery list from Sam's Club. Photo-to-cart grocery ordering: the list never gets retyped, it just gets bought.
Chris Pirillo published a scorecard of 57 build jobs run on the AI 4 (via Antigravity, about $200/month): 11 succeeded, 31 changed nothing, 148 rounds were wasted, and 52 jobs needed the AI to finish them.
This only partly worked. Worth knowing before you try it.
On a free year of the AI he has built and scheduled a stack of personal skills: a calendar manager, an expense tracker fed only by emailed invoices with his full history backfilled, a monthly utility cost comparer, and shopping re-up tracking that watches how fast he uses things like coffee beans and toilet paper and schedules the next buy. A weekly skills manager rates all the others and suggests improvements.
His 25 year old Intel Play QX3 microscope had no software that would run on a modern Mac, and he pulled it out for his grandkids. He worked with AI in stages in one morning to build a Mac app for it, and his grandkids were soon watching a seashell through it.
Tom Mulholland, 66, owner of Mulholland Grocery (Malvern, IA; team of ten; US Chamber 2026 top-100 small business), had the AI produce social ads: he sent pictures of warehouse-catalog items and got ads in ~15 minutes vs hours of typing.
Today he and the AI hacked a cheap Wish smartwatch he had abandoned for years. Now the watch talks to an AI model running locally, like a bargain basement Alexa.
Jacob runs four AI products alone. He built Tracetify, a tool that rebuilds how a competitor got its first users from twelve public data sources, in about three days, and gave the AI one instruction for the whole interface: you decide. Users later asked whether he had hired a designer.
The same studio's agent Timmy orchestrates more than 20 tools to turn blog posts into short videos: Veo generates scenes, ElevenLabs synthesises a custom voice, and automated captioning, lip sync and final assembly finish the piece. The pipeline runs inside the studio's normal the AI production setup.
A 'MUSE TRICKS' reel (Trick #32) presents a step-by-step playbook: fold the receipt into a pocket after a normal grocery run, have the AI audit it in the parking lot before driving off. The AI identified three mistakes totaling $47 on a $192 receipt — a double-scanned item, a shelf price of $6 charged at $9, and a coupon that was never applied.
The AI read a 12-page cancellation policy, found the 11:59pm cutoff, and moved the booking before the deadline passed. The original post carried a referral code.
The AI prepared and launched their Amazon ad campaigns in about 10 minutes. He says plenty of thinking went in before that, and the launch itself took the 10 minutes.
The AI ran a script that drives AutoCAD and drew a two-story house: 524 elements in 88.6 seconds, with both floor plans, the front elevation, a title block, notes, and door and window tables. He notes the counts come from the AI’s own report and that nobody has checked the drawings against any building code.
This only partly worked. Worth knowing before you try it.
Lexie Alford asked the AI to turn 6,000 saved Instagram posts into a mapped Florida bucket-list organizer. Saved-post archaeology turned into a usable travel map — the same pattern as the recipe-PDF case, applied to travel.
After a $400 rent hike, a tenant fed the AI his full lease plus three years of landlord emails; overnight it matched open repair requests to a clause barring increases while repairs are open, totalled $6,200 overcharged across 18 billing cycles, and drafted a demand letter. He later used it to coordinate 14 other units, and the landlord’s lawyer reportedly confirmed the numbers, with $91,400 at stake building-wide.
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.
Smoke from his sauna stove leaked into his home studio. He asked his AI assistant to check the AirGradient sensor in Home Assistant, it told him to leave until the air improved, and it wrote a script that polls particulate levels and pings him on Telegram once it is safe.
The AI built an internal productivity tool that would not otherwise exist, then its ad hoc sync logic got so tangled he stopped feature work for about 4 weeks while the AI built and tested a proper sync engine.
This only partly worked. Worth knowing before you try it.
A puzzle solver designed tests for whether a circulated image hid steganographic data, and used AI to help write the scripts and qualify signal and noise across JPG compression formats. He ran the tests with standard tools on every copy of the image he could find, and nothing extracted from any of them, which settled that the image itself carries no hidden payload.
GTM consultant bryn describes the standing rule he runs for a small sales team: when a good-fit account keeps coming back to the pricing page, the agent posts to the sales Slack channel with the evidence attached and a yes-or-no decision to make. The point is that reading and assembling evidence is the agent's job, not the rep's.
He says he has zero knowledge of coding. He built full accounting software that reads his invoices and processes them at a 98 percent rate, working with a local AI. It replaced the package that cost him $1,000 every year. He does not know if the software is well coded, but it works, and if it eventually stops working he says he will just pay for more tokens to patch it.
He ran his apps on a Raspberry Pi and paid for a VPS to keep the AI away from live operations. The AI worked out the spec for a PC that could run two virtual machines, researched eBay and found a good deal, and he followed its instructions, mostly by pasting commands on his phone, to set up both machines and move everything over.
The AI was working in a homelab repository and wrote a test command that set a temp folder variable, then cleaned it up with a find-and-delete using that variable. An earlier step in the chain failed, so the variable was never set. Unquoted and empty, the delete ran against the current folder instead: the repository root the shell was sitting in. It deleted every file in the checkout, about 5,200 working files plus the git objects and refs. Only files the sandbox mounted read-only survived.
This one went wrong. See what happened before you try it.
An angler who designs fishing games used AI Build to make Bank and Boat, a free fishing game for phone and browser. It covers day tickets on UK canals, rivers, lochs and piers, and it is live now with no sign-in.
Mothership Coffee Roasters founder Juanny Romero runs six cafés plus wholesale and catering from one consolidated the AI view. Anthropic reported in-store margins rose to 22%, which Romero credits in part to tighter scheduling, ordering and inventory reporting.
The Verge's Emma Roth linked her Amazon account and asked the AI to pick workout tops by size, style and color. Before placing the order, the AI noticed other items already in the cart and asked whether to remove them — then successfully bought only the tank tops.
Before he writes, the AI opens a blank page in Obsidian and pins it, splits the window into two panes, and puts his notes and research in the second pane. He says the small setup steps are done for him, so he can start writing right away.
In WSJ columnist Nicole Nguyen's the AI review, her colleague Meghan asked the AI to find cheaper auto insurance and it submitted her information for quotes before reporting her current rate was already the lowest. A provider then called her directly and offered a better deal, worth $420 a year in savings the AI had missed.
$420a year
This only partly worked. Worth knowing before you try it.
He used AI to make a Telegram channel for NHK's Documentary 72 Hours. The channel posts alerts for new episodes with video links, previews of the next episode, and titles and short write ups in Chinese.
He built a workflow that connects to their bank accounts and credit cards through Plaid and emails him and his wife a budget update three times a week, showing how they are tracking against the budget, with spending analysis and suggestions for cheaper alternatives.
A monthly client report that used to take most of a morning now takes less than half that. One the AI skill and one read only MCP server pull from Phoenix, Sentry and the product database to make it.
He used AI to move his hosting service. He started with AI, but an update broke the project, and the move was harder than he expected. With the AI's help he got through it, with a little left to finish.
This only partly worked. Worth knowing before you try it.
Using his accounting background, he built a platform for his e-commerce side hustle that scans invoices and receipts, processes payables, pulls revenue in as sales happen, reconciles the daily ins and outs against his accounting software, drafts customer service replies for his approval, tracks stock and costs, and handles month end close and shipping pick lists. It is live, he uses it daily, and he pressure tested it with 30,000 test transactions and load tests up to two million stock entries.
He found a 24 chapter site on CAD history that would once have taken days of fiddling to convert by hand. Working with GPT for a couple of hours, he turned it into a clean offline EPUB for his phone, with linked footnotes and captioned illustrations.
Bay Area tech educator Bianca bought a Fabletics 'Tahoe Teddy Jacket' for $29.99 and was silently enrolled in a $69.95/month VIP membership with no email notice. After being charged, she asked the AI to cancel it.
Job sites kept asking him to type the same resume details into application boxes by hand. He used AI to move the information from his resume into the boxes for him.
He built three Obsidian plugins for his home Draw Steel game. Over one weekend, with AI, he got them ready for public use and anyone can start using all three now.
He had the AI start his tax work this year and it caught that his property manager had double billed him for invoices on an investment property. The catch saved him $6,000, and the token cost of finding it was about $6.
He set up an AI that wakes up when a new the AI release lands. It reads the changelog, picks out only the parts he cares about, and sends him a DM when there is something of interest.
The Product Channel had the AI track a $744 ResMed AirSense 11 purchase three times daily through its 30-day price-guarantee window. It found a $634 price and — after approval of the drafted claim — secured a $121 refund; a second $81.40 claim remains under review.
Built a discontinued-products memorial site (about 239 products) with an AI research pipeline and a human review gate, which caught an AI claim that a snack had been revived when it hadn't; two weeks post-launch Google is routing discontinuation queries to its pages.
His insurer would not pay for a damaged well on one of his gas leases, and two insurance law firms told him he had no case. The AI read nearly 230 pages of contracts, letters and legal opinions, found his policy kept older, more generous terms from 2015, rewrote the claim, filed it by email, and he got the money.
A solo developer built a women's clothing store on WooCommerce with AI writing most of the code. The store matches the client's design exactly, syncs stock with her 1C system in real time, and cost nothing in platform licence fees.
He found the GPX file from his 2013 Mongol Rally, 437 kilobytes with 2,226 track points and 98 check in messages. After a flat map first to check the data, three rounds of prompting with AI turned it into an interactive 3D globe with every message pinned where it was sent and 99 photos placed along the 11,138 km route.
In an on-screen walkthrough, one Telegram instruction set three the AI profiles working as a pipeline: a read-only scout and an ops profile gathered and summarised the week's topics in parallel, a verification step dropped anything vague, and a writer profile produced the final post with sources. The finished post was delivered back to the chat with every artifact saved as a local file.
The AI wrote the tooling to annotate historic map tiles and a CV model ran across the rest for an exhaustive study of unlisted formal gardens in Ireland; a decade-planned project done in a weekend.
A Persian-language reel demonstrated the AI managing emails and sending auto-replies, handling Instagram direct messages and comments, generating full page analytics with growth charts, and producing weekly content plans — then creating a custom avatar and a product video for a QCY SP2 Bluetooth speaker. It connected to Gmail and Instagram.
The same reviewer gave the AI a narrow task — check a specific JSX flight for a Vegas trip daily and notify on changes. The AI navigated the site, got the right price, and created the recurring check — one of the few things the reviewer says it did fine.
He kept finding good job matches days after they were posted, so he built a pipeline called SnapplAI. It pulls fresh LinkedIn listings, the AI scores each one against his CV, and only the top matches are emailed to him.
This only partly worked. Worth knowing before you try it.
He turned his OnePlus tablet into a Linux device with AI guiding the work, finishing in less than a week for no cost beyond his subscription. Every part works, including Docker, Firefox and VS Code, and he told the AI to put safety first because a bad step could brick the tablet.
Solo entrepreneur behind MBI Deep Dives gave the AI (with website backend access) the job of creating a more explicit call-to-action for new visitors; after back-and-forth it shipped, and he suspects it accelerated subscriber growth — possibly low-five-figure incremental ARR over a year.
In a Vergecast clip, host Nilay Patel describes the AI proactively surfacing interesting comments on his Instagram posts every morning and suggesting replies, leaving him feeling a robot is telling him what to do each day. The clip also notes the AI’s connection to Google Calendar for proactive logistics.
He set up a pipeline he wrote himself. Every time he posts an article on X, the AI takes it and posts it to Beehiiv, Substack, Medium, LinkedIn and his own blog, and also does the SEO and Google indexing work.
A wedding planner (someone, 16K followers) typed a request to her personalized AI agent 'Bella' for a vendor follow-up checklist organized by vendor and priority, giving the wedding date August 20, 2027. The AI generated a comprehensive spreadsheet titled 'Wedding Vendor Checklist' with tabs: 'Setup & Legend,' a 140-task 'Vendor Checklist' across 17 vendor categories (Venue, Catering & Bar, Florist, Photography), a 'Vendor Dashboard' summarizing task status, and a 'Countdown Plan' with