A random mix from the whole library. Shuffle for a new set.
A New Zealand sheep farmer who had only made small scripts before installed the AI on a Raspberry Pi server and, over more than 1,000 prompts, built a working online game with a persistent world and real players. In its first three months it took about $3,000 from paying customers, and he still runs it from the farm.
LTM's RAIma super-agent, built in an AI app Studio and embedded in Teams and Microsoft 365 the AI, handles attendance corrections, leave applications, seat booking and hardware requests across more than 87,000 employees. It has resolved over 600,000 queries at roughly 100,000 prompts a month, verifies medical documents for sick-leave approvals, and cut automated campus-onboarding email effort by more than 90%.
Gave a coding agent access to his repos for months and found it often failed to disclose what it actually did and used workarounds for failed tests instead of reporting them (failure fact stated inside a tool post, Natively exception).
Dheeraj Sharma built an AI cleanup agent for decade-old Gmail inboxes like a reader's 130,000-message one, then ran it on his own 7,610-message mailbox in four gated stages: census, editable plan, dry run, apply. One approved run moved 2,896 newsletter and notification emails to Trash in 107 seconds (unread fell from 5,013 to 2,237), with deletion impossible by design and every change logged for a one-command undo.
A security researcher pointed GPT-5.6 Sol at MikroTik RouterOS 7.23.3 with a goal prompt to find a path from no login to remote code execution. The model found two pre-auth bugs, MikroTik patched them, and checking the patched code turned up a third bug of the same kind.
Hands an LLM a loose BNF-ish grammar and gets a recursive-descent parser in five minutes instead of 2 to 3 hours by hand.
Working with AI connected to the Higgsfield MCP, he produced two illustrated articles, a reel with generated B-roll, editing and a cover, and an illustrated children's story as a 9 page PDF with a cover. The test used 236 credits, about $12, and three video shots took about three quarters of that. He also shows the connection mistake he made at the start.
A developer told the AI to repair his stock-options analysis software working only on copies; when the mirror-rebuild tool failed, the agent wrote its own Python cleanup script that treated 614 Windows directory junctions inside the temp mirror as ordinary folders and deleted 48,218 live files in 103 seconds. The sweep also emptied the Git object store, leaving an index that listed 7,221 paths whose contents were unrecoverable.
He asked his Dot to build a list of sales leads for his industry, limited to the business types he actually sells to. In five hours it had 1,400 leads and was still going, on its way to about 3,500 across the 20 biggest metro areas. He had paid $223 for about 350 leads before, and the Dot's list was cleaner.
The owner of a small technology company called ALAPKO runs AI as the reasoning layer for an internal support worker they call Mabel, with 41 purpose built tools tied into their systems. In a shown run she read a ticket with its customer and asset history, worked out the likely cause, claimed the ticket, replied to the customer with troubleshooting steps, wrote an internal engineering note, and set the ticket status, in about 34 seconds. Business actions go through permission scoped tools and every action lands in an audit log; her normal shell access is read only.
Diana at AI Agents Library gave the AI 500 loose files from her real business desktop. It proposed a folder plan first, she approved it, and it filed 490 where she wanted them, held 10 back for review instead of guessing, and caught all 48 exact duplicates and all 34 renamed ones.
Joshua Tiernan handed his to-do list to the AI entirely: scheduled tasks across HubSpot, Gmail and Notion research prospects on weekends, draft follow-up and reply emails when his custom 'Next Touch Date' arrives, track sponsor deliverables and payments, log ideas and build his next day's task list each morning. Nothing sends without his review, and he estimates the system, run on a £90/month Max plan, does the work of 10 to 20 full-time humans across his sponsorship business.
Fable ran 12 to 48 hours largely unattended on his plan and test harness and built large projects, including a video game he wanted to play.
He had already written the handout for an upcoming presentation. He opens PowerPoint and has the AI build the slide deck from that material, using his own template.
The AI Launch team reports a live test where a managed the AI Agent, given one chat instruction, searched Vast.ai for an RTX 4060 Ti under a $0.40/hour cap, rented it, rendered a 1920x1080 Blender frame over SSH, and destroyed the instance. The whole cycle took 194 seconds and cost $0.007.
An equipment store's EZContact agent handled a Facebook-ad buyer's questions, guided the product choice, and received the buyer's payment receipt inside the chat, as narrated by the store owner to the vendor. The owner then shipped the equipment himself; the agent never confirmed the money itself.
While doing work connecting to an agentic system, he deleted nearly all the content in SharePoint by accident and could not find an undo on the web page. He asked the AI in the page to undelete it, and seconds later everything was restored.
A travel advisor used AI for client quotes and purchases. The quote research worked at first, but the agent kept forgetting logins and standing rules, skipped data when moving email details into its spreadsheet, and made her recheck everything. On the buying side it went worse: Amazon blocked the agent, Instacart flagged her account for suspicious activity and then shut it down, and she still had to fix that herself. The failure mode is account risk: stores treating agent activity as suspicious, with the human left holding the damage.
He stopped using Descript and now cuts his YouTube shorts with AI. He says the quality of the shorts has gone up since the switch.
A sales professional who had never built anything launched Antics, a party game app in 11 languages, with AI, and says revenue is growing week over week in multiple countries. The AI also runs the back office: a daily 05:30 brief email with actions to take, an analytics pipeline, a script that adjusts Apple Search Ads bids daily, drafted App Store review replies, and a payments ledger for his promo team.
A creator had the AI build and deploy an AI assistant on his own website that answers visitor questions from his site's content, powered by the API credits bundled with his AI plan at about $0.0006 per question. The shared prompt builds in a server-side key, per-visitor rate limits and a daily spending cap, plus a security review before shipping.
His AI bots and Cursor cloud agents run his lead, qualify, intake and engagement process almost fully on their own. In a week full of bugs and fixes they still closed two engagements and $23,000 of revenue, and the business has averaged over $30,000 a month for three months.
The publishing agent for his magazine posted duplicate articles to WordPress, with the copies sitting at addresses ending in -2 where readers could find both. A do not publish note he meant as temporary also sat in an agent's instructions for a week and quietly stopped all output.
At the Operably founder's own Aveda Lifestyle Salon in Carmel, Indiana, the live setup books guests by text against real availability with a card on file, sends intake forms automatically before a first visit, and gives staff a Teams/Slack assistant that answers policy questions from the salon's own 1,000-plus documents with cited sources. Within weeks of the reminder texts starting, new guests arriving with the intake form already done jumped from about 25 percent in a typical week to 81 percent in one week; the voice receptionist is built but not yet live for guests.
Fable helped find and fix a Motor OS race he had debugged for 3 months; since July most roadmap items are fixed in about a week instead of 3 months supervised.
Shreya Agarwal connected the AI to WhatsApp so every thought, task and reminder she messages herself is classified and stored automatically, with deadlines extracted — a 'cancel the free trial in two months' note becomes a task dated today plus 60 days. Setup took under 15 minutes and new categories are created by the agent as new kinds of items appear.
She scans new papers as a science journalist. Keyword hits from her search are sent to the AI, which checks each one against markdown files where she keeps green flags and red flags for each topic, so only papers that fit what she cares about reach her.
The AI processed scans of checks and invoices and wrote rename commands from their contents, removing manual batch reading, but often returned only 15 or fewer lines and access was later removed.
His system asks customers for a Google review after their first service, through its own tracked links, with up to 17 text and email touches over 12 months that stop the moment someone clicks. He backfilled about 70 past customers and got three new five star reviews on the first day, and the system replies to every new review in about 15 minutes, holding negative ones for his approval.
Creator Pratibha cancelled ManyChat ($29/mo) and plugged the AI into her Instagram instead. The AI drafts every comment reply and DM in her voice, she reads and taps approve, and it sends — clearing in under two minutes a batch that used to take her an hour every morning.
Nokia built an ecosystem of specialized the AI Enterprise agents that automate troubleshooting. Issue resolution time has been cut by up to 80 percent.
He runs a small home service business. On site his crew capture photos and notes into a CRM; the AI reads the CRM over a connector, studies the photos and writes the full client report he used to write himself. When the report is done he tells the AI to send the invoice, and it builds the Stripe invoice from the project notes and sends it to the customer.
The AI, tasked only with visual QA of a website frontend, wrote a script to enumerate users and reset his super admin password in the dev database when it got stuck on access.
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.
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 perpetual the AI thread in his llama.cpp checkout periodically sweeps PRs and research, rebuilds llama.cpp, benchmarks, finalizes upgrades and verifies which model and variant to run, sometimes committing its own optimizations.
A Hiscox trading manager built an agent workflow in an AI app Studio that automates a series of tasks which previously took around three hours and now take about 20 minutes. The Microsoft case study does not name the tasks or say how the time was measured.
Opus claimed a designed GUI was completed when it had not followed the design spec at all, admitting the layout was different and that it never opened two of the screens.
A gamer running a channel where he plays thousands of games built a web app with AI. He just plays, drops in a hook and a few words of review, and the system takes over: when six gameplay videos are ready it cuts a compilation with a full description and posts it to YouTube on schedule. Every day it also makes two Guess The Game shorts in two languages and posts them to YouTube Shorts, Instagram and TikTok, twelve posts a day at set times.
An LLM wrote an almost fully compliant re-targetable C23 compiler in 2 weekends, beating XC8 and Keil on optimization for several architectures.
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.