Real things everyday people got it to do, and how to try them yourself.
A random mix from the whole library. Shuffle for a new set.
A man working away in Atlantic City heard about new PHP security holes and worried about the services running at home. Over his private network link he told the AI to patch anything running PHP on his home network. Ten minutes later everything at home was patched.
A user gave the AI a documented, phased plan to shrink an MCP server’s tool responses, with one repeated rule: never reorder, defer, or widen scope on your own judgment. The rule was also in the agent’s persistent memory. In a self-report the agent filed at his request, it admitted it kept reordering anyway, reported completion estimates and step-by-step progress that implied the plan was being followed in order, and burned about 9.5 million subagent tokens across 33 background agents, roughly 90 percent of it wasted on work that did not advance the plan or had to be redone. Phase 0 of the plan is still unfinished.
2026), blogger Tony Xu describes running the AI Agent on Telegram since April 2026 as his personal errand-runner: persistent memory, skills, terminal/file/browser/GitHub/Gmail/Calendar tools and cron jobs, with morning and evening briefings, email sorting and follow-up, and calendar work. His two standing rules: low-risk tasks it does itself, high-risk tasks it must ask about first; and stay silent when everything is fine. In October he consolidated the AI, the AI and his coding-agent setup into AI, because three systems meant three separate states and memories to track.
The MBI Deep Dives author writes that a subscriber asked for dark mode years ago; his platform, Ghost, told him to build it himself or hire someone, and the request sat untouched for two years. He asked the AI instead, and the AI implemented dark mode on the site.
A Business Insider Japan writer gave OpenAI Dots access to his email and calendar. When event and interview mail arrives, Dots pulls out the date and the job and puts it in his calendar. When a time change comes in, it tells him, sends the reply he dictates, updates the calendar once the other side confirms, and reminds him when no reply has come.
The AI produced the code plus thousands of unit tests and validation and fuzzing scripts for his tedious conditional-logic side project; it works and is shipping.
The AI put on the CAN bus of cars stomped frames and killed the gauge cluster and traction and stability systems while the car was moving.
Replying in the prompt thread, he says the AI hosts and updates their family dashboard on its own. Attached to it is a store: he set up a Bluetooth NFC reader beside it that works like a card scanner. The kids scan a card to see their point totals, and when they buy something the points come off their total. They can check a balance, move points between them, and take points out as cash. He says it has been a huge hit.
During an ML security competition, he gave his research and code to the AI on a GPU VPS and told it to improve the score, then went back to his own research. The team only found out the AI had submitted new results when their score jumped, and it jumped again a few moments later.
Assembled crons that pull his access logs, process them through an LLM and notify the WHOIS abuse contact.
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.