Harvard physicist Matthew Schwartz built BootLoops, an open-source harness that points the AI at exact, checkable calculations, and ran it for three months across roughly 400 candidate problems: the result was 36 manuscripts in 18 fields with 19 coauthors, including 15 Feynman integrals never computed before, Watson's 1939 'final problem' solved, and a forest model showing Panama's Barro Colorado species mix changes 4.5 times faster than neutral theory allows. The AI reproduced results from his own paper in about 20 minutes that had taken him weeks to code.
Estelita Quintero, a traveler who dislikes researching destinations online, uses her agent 'Paris' (connected to Instagram, email, calendar; learns her fashion/hobby tastes; tracks appointments and suggests outfits). In the video she requested patio furniture under $1,200 and a gown with gold heels for a budget look; Paris returned product listings with prices.
Leo Batistella (Portuguese-language creator) showed a chat where his AI listed his itinerary — Oct 9 Bali to Shanghai, Oct 10 Shanghai to Singapore (early morning), Oct 11 connecting in Singapore to Toronto without leaving the airport — then questioned whether the Singapore-Toronto leg was Oct 11 or Nov 11, noting his Toronto hotel is booked Nov 11-15. Without being asked, the agent checked his ticket and confirmed the flight is in October, not November.
His landlord had been passing property tax on to him incorrectly. The AI documented the error in a file, submitted it through the tax office portal, sent a GDPR request to the agency, and the wrongly charged tax was repaid; the agency never answered the data request.
A home-lab owner reported on X (via WhatAgentsCanDo, Oct 19) that, still in bed, he sent the AI a link to a post containing a setup recipe for a new DeepSeek model (DeepSeek-V4.1-Flash-1017, EXL3) across two NVIDIA DGX Spark machines and asked it to do the setup. The AI was not installed on the Sparks, but they shared a network with its computer, and everything was ready to test by the time he got up; he says he does not know exactly how the bot reached the machines.
A forum user gave the AI the days his family could go to a Christmas season show that runs all through December. The AI went through the ticket site day by day, found the seats, and ranked the options by price and seat placement, which he says helped the family pick the best tickets.
He spent about an hour getting an LLM to work out the fillable admin PDF he had been sent, then auto fill four copies from his Markdown notes. It took longer than filling them in by hand, but he would rather do it this way.
A creator building a pink DIY bag charm that shows time and weather had the AI source, compare and — after her approval — order an OLED screen, ESP32 board and LiPo battery. The finished charm is shown working on her handbag.
ThursdAI host Alex reported that his AI 'Wilfred', newly flipped to the AI Wonder on his paid plan, found the New York podcast studio he recorded the show from. The AI then emailed Ezra at Virgo Studios with his requirements.
Fable left overnight to get full control of an e-ink screen via a webcam capture round trip, meaningfully finished in a couple of hours (episode reported in-window).
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.
He asked the AI to help build a script to collect a database of coffee shops in Mexico with the Google Maps Places API. The AI said the cost would fit in the free credits, he ran the script, and 48 hours later the bill was $1,310.05.
$1,310at stake
This one went wrong. See what happened before you try it.
Three family members, three nationalities, one move to Brazil. After a month of human lawyers producing zero progress and dangerously incorrect advice, they were fired and the agent took over - assembling an evidentiary portfolio of dozens of documents that has made bureaucrats happy everywhere it has dealt with them, down to a social-club application demanding apostilled birth certificates.
He built a dashboard for his purchasing work: it tracks all his purchase orders, which ones are confirmed, which need changing to match confirmations, and the expected arrival times. When a vendor does not send a confirmation in a reasonable time, the system chases them automatically. He says his metrics are better and the job is less stressful.
Andrew plays casual online chess against a friend he is dead even with, and was about to buy a Chess.com membership for unlimited game reviews. Instead he gave the AI his Chess.com account; it automatically pulls and reviews his game log every day, and the premium upsell lost its reason to exist.
The road to a takeoff spot they fly from was not on OpenStreetMap, though satellite images showed the track. They built a tool called osmctl, and the AI drove it and mapped the road.
He had a bunch of Amazon returns pending. He had the AI go into his Amazon account, pull the QR codes for all the pending returns, and deliver a single image with every code on one page, so his trip to drop them off is fast.
Same reel as the recipes row: Aggie Lal's AI agent 'Munchkin' summarized her ~14,000 saved Instagram posts and generated a second PDF, 'Parenting Tips', covering postpartum recovery, mom fitness, and breastfeeding nutrition.
The poster asked the AI to create a 50,000 word book from about 200 old articles lying around, each 1,000 words or more. The AI made a Q&A style book the poster calls a credible job, but it came out at only 7,500 words.
This only partly worked. Worth knowing before you try it.
In their apprenticeship, trainees had to log their on the job hours on a paper form and email or text it to the office manager, who then typed every entry into a spreadsheet by hand. He built an app so apprentices can submit their hours directly. Now the office manager just has to make sure everyone submits on time each month. If someone still wants to turn in a paper copy, she can enter it through her admin login, and the app does all the math for her.
He runs the AI on his PC with ha-mcp on the same machine, so the AI can reach into Home Assistant and make the changes he asks for. He keeps it on his own PC only, and says a new dashboard is done as soon as he describes it.
Creator Wendell Yu says the AI read his 17 years of Facebook posts and wrote a profile report of his life from them. The segment is the first of five first-hand cases in the video.
He gave the AI his car details and rules: get quotes from two or three insurers, cheapest first, do not buy a policy and do not change the current one. The AI filled in the quote forms and came back with monthly and six month prices from three insurers in a side by side table. Geico would not complete the request at all. He says it saved him time filling in forms.
This only partly worked. Worth knowing before you try it.
He had a rough script that launched internet radio stations on his Linux desktop. The AI made it safe and finished: it checks the saved player, searches the Radio Browser database, adds tests, a manual and a README, and pushed it to GitHub.
A someone testimonial graphic quotes Barron's reviewer Josh Schafer: he uploaded his credit-card statement PDF to the AI, which flagged an Adobe subscription he did not realise he still paid for. Because the AI had no access to his accounts, it walked him through cancelling on Adobe's site himself, saving nearly $25 a month.
He forgot to charge the batteries for his road bike's electronic shifting before a ride last week. After a weekend of back and forth with AI, he now has a dashboard that can alert him when things need to be charged.