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.
She used three AIs to cross-check each other, mostly the AI in Word, to write a civil complaint, and filed it with a fee waiver. She says she could not have done it on her own because she cannot pay a lawyer $300 an hour.
He had the AI make an app that controls his Mac fully from his iPad. It took two sentences and two follow up prompts, under two hours of work, and the app even built its own recorder to check its streaming sessions.
XAI's the AI Voice 3.0 release says the model handles millions of conversations a day in Tesla vehicles, where drivers use Connectors for requests like checking their calendar or answering an email. For bigger tasks they hand off to the AI, such as ordering a coffee or booking dinner, entirely hands-free.
Unite.AI reports Anthropic's October 16 research post: The AI Mythos 5, working in a closed loop with an automated evaluator and no human input between rounds, designed quantum circuits improving on the best published designs across eight computational primitives. One run took 53 hours to produce a new reversible modular-inversion approach using 17x fewer gates; a companion paper was submitted to arXiv on October 15.
He built a Homebridge plugin with AI that lets him turn his server on and off from Apple HomeKit for free. He says he does not know the first thing about coding.
The AI built a complex quiz web app for himself and friends with multiple question types, screen sync, media uploads and scoring, in about a week of evenings (build reported in-window; the comment's complaint about current models left aside).
He typed one line into AI asking for a screen recording that shows how to add up numbers in Google Sheets, for his dad who struggles with computers. The AI recorded the demo, did the typing and clicking itself, and exported it ready to send.
Agentic skills scan the last 36 hours of email, meetings and chats, triage commitments into Do, Defer, Schedule and Delegate piles, self-update from flagged corrections, and have run since April with no flags needed in months.
A developer gave the AI a token for his code repo and set a two hour cycle. The AI checks the issues page, makes a branch for each issue, writes the fix or change, and opens a pull request. Merges rebuild the app and serve it on his home network, and the AI texts him on Telegram when it has a question or when work is done. He uses the app daily and files features as issues, which are handled within a couple of hours.
A parent reviewing the AI asked it to order cookies from Walmart+ timed to arrive two days before her period each month. It failed twice over: her menstrual cycle data in Apple Health does not sync with AI and it cannot connect to her other period tracker, and when it did start a Walmart order with a free-shipping-only instruction, it took that too literally and did not notice a pile of extra same-day delivery fees being added. She told it to cancel the task and not place the order.
Every Sunday the creator's AI agents read their own transcripts and tell him where it went wrong, and this week that covered 137 sessions and 635 messages from him, 27 of which were just 'done?' or 'update?'. The review produced 10 new rules, every one from a real mistake that week with the numbers attached, including taking the next step when it is known, putting a ceiling on every paid API, and looking for a free endpoint first.
Lauren El-Dahdah, a Houston wellness/lifestyle creator overwhelmed in September, tested the AI via her agent 'Mimi' (beachy avatar). For her son Leo's 2nd birthday, Mimi created a six-page printable 'Leo's Born 2 Ball Party Plan' (sports theme, Sept 26): countdown timeline, guest communication templates, daily task checklists, shopping lists organized by zone, and a day-of run-of-show.
An LLM crunched for 36 hours on a project he had wanted for 5 years; checked and ready to deploy, work he estimates would have taken him about 9 months.
The AI completed and filed the user's resident tax report through Korea's HomeTax system.
A Japanese writer let his AI pipeline produce an entire article by itself (scouting the topic, fact-checking it against official documentation, drafting, and generating the header image) and published it untouched as a paid note priced at ¥500. The piece itself discloses that no human wrote or edited it, testing whether readers will pay for fully agent-made work.
An agent researched available tools, built an eval set, and tested tools against each other to pick the set used for his singing app, cutting build time from months or a year to about 10 days.
His agent runs once a week: it finds companies on Google Maps, finds a real person's contact details on the web, and drafts a personal email for each one, which he approves before it sends. Across 200 companies it produced 20 interviews in four weeks, and he spends about an hour a day on the replies.
The AI is his gym trainer: he reports his sets and reps as he does them, and later that day Astra logs them into an Excel tracker, then uses the updated tracker to refresh an HTML app it built that shows his progress over time and his missed workouts.
He had the AI answer a Slack notification from a Datadog monitor. The AI checked the issue through Datadog MCP, found the right Slack group to ping, and sent a summary of what needed fixing.
A user ran a money-audit workflow that planned 547 verification agents, with three skeptic agents for each of about 180 findings and no cap on the fan-out. About 180 agents ran, using roughly 36.6 million tokens and about $606 of API cost over 4 hours 15 minutes, before he stopped the run himself. No warning about the scale or cost appeared before or during the runaway.
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.