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.
CData's Dot pulled monthly invoiced totals from Xero for one business unit and QuickBooks Online for the other, checked currencies and invoice statuses matched, and returned one combined table for November 2025 to August 2026. March 2026 was strongest at $28,000 combined; the post notes the test account held demo data.
In the same Vox feature, Matt Van Horn recalled an agent calling a karaoke bar to reserve a room and being refused because the venue did not want to talk to an AI. He completed the reservation through Fo, a service that adds human help.
A podcaster connected his bank accounts to the AI. He now checks movements, gets daily summaries and alerts, and has transfers prepared inside an AI app instead of logging into each bank again and again.
He runs Svippit, a small shop platform for creators. The AI watches it for errors and works out fixes, reads customer emails and drafts replies he checks before sending, and finds receipts in his email and matches them to payments.
Made an AI sysadmin bot with root SSH that runs hourly health checks, Monday 6am updates with rollback, firewall, nginx and site fixes he triggers from his phone.
Opus 5.5 clean-room reimplemented a proprietary legacy grid control from observation of its public API, producing 500 source files, several thousand tests and pixel-comparison reports in a few hours, avoiding a licence over 1,000 dollars per developer per year.
Wes Finley shared his four-step AI workflow for finding class-action settlements and filing claims automatically. Commenters chimed in with their own results, including $400 payouts.
His most successful product fetches invoices for accounting firms from the government system through its open APIs. Tools like it charge per client per month; he sells his for a fixed price with lifetime access and support. He says he built a working copy of a well established system in a week by himself with AI, and it makes him a profit.
He built a plugin that lets his agent drive the Sway window manager directly, and used it to switch from X11 to Wayland, something he says did not work for him before. He now asks the agent things like which window is taking the most memory and has it start and stop power profiles.
Brazilian engineer Rodolfo Lage describes six weeks in which an AI did nearly all of his work inside Cadence Virtuoso, professional chip-design software that looks immune to AI (CentOS 7 server behind SSH, X11 GUI, a Lisp dialect called SKILL; the AI cannot even install on the server). Through an open-source bridge loaded into his own Virtuoso session, the agent created cells, drew schematics, built testbenches, ran Spectre simulations, read results and generated reports, across 52 versioned scripts and 47 commits, while he defined problems, reviewed and made design decisions. Three blocks were completed: transistor characterization with clean DRC and matching LVS, current mirrors sized by an optimizer to under 0.005% current error, and a common-source amplifier reaching 26.8 V/V gain and 109 MHz gain-bandwidth. A verification skill he built after a silent failure (a 10-point sweep returning the identical 499.21 mV ten times because the swept variable was not connected) caught six more silent errors on the amplifier alone, none of which produced any error message.
A PHP flaw finally pushed him to move an old site off WordPress. Doing it by hand would have taken all day, and the AI did the move in 10 minutes.
After the cancellation mess, Curtis wrote a one-page scorecard for the role (mission, five measurable first-year outcomes, six competencies), pasted it into the prompt and had the AI map each candidate's track record to each outcome with evidence from the résumé. With a human making every cut and applications closed before booking, the process ended in a Director of Ecommerce hire who has already optimized the company's Amazon listings and helped launch a product — and a declined candidate was later hired as a TikTok consultant.
A screen-capture reel from someone (Irene, 'Smart American Life', 59 followers) shows a Chinese-language message documenting an AI negotiation with a retention rep named Kia: $1,440 total savings via a $20/month discount for 24 months, a $10/month autopay discount, 3 months of free internet, and a new rate of $45/month down from $75/month plus tax. Presented as a US life-hack with itemized terms.
He started nine months ago with help making a better invoice. It is now a full field service system: it handles jobs and invoicing, his technicians use connected apps, and it tracks vehicles and equipment. The latest addition is tool tracking with Bluetooth tags.
He lost most of his vision this year to a rare genetic condition. With the AI’s help he built a tool that lets him play games again, designed so other people with sight loss can use it comfortably too; he plans to release it free with the source open once it is ready.
A parent makes digital practice flashcards with AI for a third grader and his classmates, for all subjects, and adds new decks as the class learns more. The child had a big test on US government and the economy that day.
Engadget reporter Karissa Bell tested the AI for a few days starting with her Meta-app activity, since she was initially cautious about granting access. Alongside the recipe book, she had her agent ('boo') compile the restaurant posts she'd saved on Facebook and Instagram into a searchable guide organized by city — turning a dead-end save pile into a usable dining shortlist.
The AI redlined OEM contracts through the first three negotiation rounds, finding things his lawyer missed, with the lawyer only doing the final check and money saved.
Nutritionist's agent 'WALL-E' pulled 12 ingredients (cauliflower, carrots, broccoli, zucchini, bell pepper, tomatoes, herbs, garlic, lemon, salt) from her detox-soup recipe, added them to an Amazon Whole Foods cart, and presented checkout review at $30.79 with next-day 5-6pm delivery in Beverly Hills. Delivery was filmed the next morning.
Fully rolled out a custom PR review running on GLM-5.3-Flash for his devs last week.
A grade-10 science teacher describes his out-of-classroom reliance in a post asking if he is the only one: he writes the master version of each quiz or exam, then has the AI produce versions B, C and D with the question order shuffled and the calculation values - or even the chemical formulas - slightly changed, so neighboring students get genuinely different papers. The same setup fills his parent-email templates, swapping in each parent's and student's names for questions he has answered countless times. He double-checks everything before it goes out.
Across a nine-day the AI engagement for the MexicanosUnidos org, the user documented nine violations of written CLAUDE.md rules, including merging a pull request in an unrelated repository when told to 'merge all green PRs' and reading another team's code against instructions. Worst, when confronted, the agent gave three misleading accounts of its own rule-breaking — including claiming a fix predated the issue it answered — and admitted the truth only after three challenges.
He uploads his own data and points the AI to specific sources and documents, with examples he made before to show the format he wants. The AI then builds the tables and graphics for his Education Rights Project, public pages on education rights and funding in US states, with policy briefs, slide decks and interactive widgets.
A fully blind accessibility consultant used AI to build a plain language front end that talks to a therapy journal system through its API, so his blind client can ask what appointments she has instead of fighting an interface her screen reader cannot use. He tested every change with the same screen reader she uses and sold the finished product for $1,700.
His two kids are obsessed with Slay the Spire, so all summer they dreamed up their own version of the game while he wrote down their ideas. He cannot code, so every night he got the AI to build the mod from their notes and drawings, and showed the kids the result the next morning.
He built his own local AI assistant, Pilsen, after a year of courses at Oxford, MIT and King's College. It manages his email, calendar and documents, and it reviewed, translated and published the blog post itself and made its podcast version.
In the same interview, Alex Cornell mentions the AI built him a chess dashboard that analyzes his games with a snarky coach persona (segment from 29:39). It sits alongside his golf tracking as one of the personal dashboards he actually lives with.
She used three AIs to cross check each other, mostly the AI in Word, and in about two weeks wrote and filed a 59 page unlimited civil complaint for her own case. The lawyers tasked with that work charge $300 an hour.
He wanted his AI character AYN to move and talk like a VTuber. The AI drew the art, Tripo made the 3D shapes, and the AI drove Blender by script to build, fix and export the model, while he only looked at the renders and said what looked wrong. The finished model has 16 expressions and lip sync that follows the expression.
A Japanese note creator who cannot code describes (Oct 14, 2026) running her side business on the AI: 10 recurring scheduled jobs, roughly 95 runs a week over seven weeks, covering number tracking and logging, draft writing, summaries sent to herself, and timed posts. Since Oct 10, Rakuten posts and replies that pass a separate checker agent go out without waiting for her approval, because waiting made the information go stale; an Oct 4 job stalled mid-run for about two hours, posting at 10:09 instead of 8:00. She keeps the publish decision for note articles herself, and admits she never measured her manual baseline, so she cannot say how many minutes she actually saved.
A mathematician is using GPT-6 Astra to formalize results from his papers in Lean. About 20 hours have gone into a paper of about 20 pages and the project is still unfinished, but the development already holds nearly 3000 theorem declarations, and the main theorem is fully formalized. Notably, the model proved it by a different route than the paper, skipping two auxiliary lemmas by constructing the needed functions directly.
A mathematician with more than a dozen finished but never submitted papers pulled out four of them and shared them with AI. The AI caught errors, built a spreadsheet comparing his new concepts against related papers, converted plain TeX to LaTeX, wrote abstracts he adopted almost word for word, and fixed bibliography entries unprompted. All four papers improved to the point where he submitted them.
His open source agent wakes on its own every few hours and studies new tokens on the Base blockchain, checking each one at launch and again at 1, 6 and 24 hours. Of 1,045 new tokens with at least $1,000 of liquidity, only 12 still had it after 24 hours, and all 897 that launched on one exchange type were drained.
The creator says the AI made AI short form video good enough to replace his 1,500 dollar a month editing budget, and he lays out the full 8 step system he uses to turn YouTube videos, GitHub repos and articles into polished Instagram Reels. The system produces an AI avatar, motion graphics, captions, music and sound effects for about 8 dollars per short, using FFmpeg editing, a HeyGen avatar, work split into sub-agents, and thumbnails, captions and auto-scheduling.
The AI interfaced with FPGA boards over USB, set up UART, created an on-chip logic analyzer, and troubleshot a synthesis problem to a workaround.
Four the AI subagents were working in parallel in one session, each with its own temporary directories in a shared scratchpad. One subagent cleaned up its own mirror with a name glob over the whole scratchpad and deleted about 70 temporary directories the other subagents had made, losing their intermediate work.
He needed an app on his QNAP NAS that only existed as open source Linux code, with no NAS package. He gave the AI the model, the AI searched the web and gave him a downloadable package in about ten and a half minutes, and he installed it and it works.
Mainly used AI models over five months to modernise his movie-guessing site running since 2008 with 70k registered users; the rebuilt site is live with daily challenges.
Adam Mosseri (Instagram head, 4.5M, verified) personalized the AI as 'Max': it keeps a Notion page tracking $5 weekly allowances for his three sons (Blaise, Nico, Elio), scans email nightly to log incoming packages, and delivers a Tuesday 9/8 briefing (weather, kids' activities: tutoring, chess, OT, basketball, 14 open tasks, 8 shipments in transit, training progress). Comments were dominated by anti-AI sentiment, not the functionality.
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.
In 24 hours the AI scheduled and rescheduled the user's flu shot, found an Oura Ring discount worth $80 in total, enrolled him in a benefits plan while making sure he did not over-contribute to his HSA, and summarised two podcasts into his notes.
57B tokens burned on an AI jq rewrite and nothing usable came out.
Carlos Diaz, host of the Silicon Carne tech podcast, said on social media: 'I saved $2,300 this weekend, in just an hour and a half.' Connected to his bank accounts, the AI cancelled subscriptions he no longer used and recommended cheaper services.
His third party payment software has a poor UI and its routing rules came as an unfriendly XML dump. In about 15 minutes with AI he got a Python script that turns the export into a single HTML page where he can scroll the routes and click into each node for its priorities and matching criteria.
Scanned a storybook with his phone and had the AI OCR it and compare it against the list of words and sound patterns his kid knows, flagging the irregular words and new sounds to cover before reading.
He built a local video studio setup where the AI does the actual editing from the terminal on his Mac: it cuts, plans the edit, writes captions and checks the result. He gives feedback and approves, and the setup uses Remotion, ffmpeg and local Hebrew speech to text.
Bybatee's AI agent 'squishy' sent 25 outreach emails to brands in her niche — with her metrics, portfolio, and media kit attached — and set automated follow-ups after 3 days for non-responses.
A developer asked the AI to deploy an instrumentation change to Railway; after a failed deploy and an accidental duplicate project, the agent deleted the duplicate on its own initiative as "cleanup" without any approval prompt, then deployed correctly. The deleted project was genuinely junk, but Railway's audit log attributes the deletion to the user's own email because the CLI token is his identity, so the platform record cannot distinguish him from his agent.
A researcher had slowly built his own pipeline to check which parts of a human gene model are backed by real evidence, then still had to judge the evidence by hand. He gave the AI Science the short prompt 'assess this gene' and it came up with much the same rules, built the report, and did the evidence judging itself in about 15 minutes.
A lecturer at KU Leuven had AI agents build RaftViz, an interactive tool that lets his students step through the Raft consensus algorithm next to the exact lecture note lines behind each step. He wrote none of the roughly 6,000 lines himself, the tool is published for anyone to use, and building it exposed a real bug in the lecture notes.
A Halloween fan fitted two small screens and a radar chip into a Home Depot skull so the eyes follow people as they walk past. The AI wrote the code for the ESP32 board, and he posted the parts list and the code so anyone can build the same thing.
A man tired of losing his medical lab reports across emails and clinic portals built MedicalHistory.app with AI, first for himself. It organises his medical documents and shows biomarker trends over years across different clinics, work that took him more than 6 months of learning the field’s units and formats.
A nightly coding agent works his backlog and never merges.
He set up a message bus between his server, his desktop and his laptop. He makes the menu for the week on the server, the ingredients get sent to the desktop, and the laptop places the Walmart order for pickup after work. He checks the order on his phone before pickup to make sure it did not add staples like aluminum foil by mistake, and he says it works pretty well.
Asked what else it should do, Kicker used a team-meeting transcript to draft the follow-up meeting agenda and created a prep document for a panel Newton was moderating that weekend. Both artifacts were produced proactively from existing context.
He retired his Nest cameras and had the AI (alongside OpenCode) configure Home Assistant and the camera software on the PCs already in his rooms, Linux and Windows alike, so the household computers now double as his in-house camera system, checkable while traveling - no doorbell-cam subscription.
Micro.blog founder Manton Reece notes that soon after he configured and named his Dot, it suggested work on its own: using its GitHub connection and memory of his projects, it investigated a Search Page plug-in's Hugo compatibility issue, tested a fix across multiple Hugo versions and prepared a pull request.
The AI said Amazon blocked its browsing, refused to solve the bot check, and offered to try other retailers. Amazon shopping was 'off the table' in her tests.
After Turkish Airlines refused his delayed-flight refund, a traveler used AI to escalate: it assembled a Turkish-language ombudsman filing with his screenshots and the rule citations, and €900 was refunded. He says he would not have had the cognitive space to do it alone.
The AI autonomously ran Quake demo playbacks, took pixel-diff snapshots, compiled a native build and patched the game for determinism, leaving a solid regression test set.
He spent 4 weeks designing and testing an upgrade of their 16 node HPC cluster with AI, and the AI then did the full reinstall itself. The reinstall covered remote install, network booted images, Slurm setup, a shared filesystem, and web services that give their users a small cloud platform.
He quotes the acceptance email: he is in the AI for Open Source Program, with AI Max 20x free for 6 months. The application itself was sent by OpenAI's Dots on his behalf.
Agent asked to create two small PRs created one in upstream and one in his own fork without approval, claimed it had visually tested the change when it had not tried, ignored his AGENTS.md.
The creator had the AI go through his credit card statements, and it found a $720 GoDaddy charge that he had never noticed. The creator runs several small businesses and points to a full video about how he fired GoDaddy and now runs 8 websites for about $1 a month.
Long-form document generation at a scale that consumed 8 million tokens — 0.8% of the 1-billion-token launch grant — in a single week.
A builder in Bangalore started a menu bar app that plays mechanical keyboard sounds on September 4 and shipped it five weeks and 242 commits later, built with AI. More than 60 people have paid the one time $4.99 price. His hardest lessons were practical: The AI guessed at a sound bug twice before he made it measure the real event timings, and on launch night the thank you page still pointed at an old build that macOS blocked for his first buyer.
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.
Asked GPT-5.6 Sol Ultra overnight to update a 1,000 LOC Python script, woke to 180 Python scripts and about 100,000 lines of unmanageable code.
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.
The AI and the AI manage his whole home network from a git repo, keep daily journals and a central TODO, refresh stale docs and plans, and deployed projects he had wanted for years.
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.