A random mix from the whole library. Shuffle for a new set.
He asked his AI to edit an entire video and it did the work, from picking the takes to lowering the music in the quiet parts. The video is the third in his series about building his toolstack with an agent.
A creator had an AI design session write the build instructions for one ambient cafe BGM video, under standing rules to check before sending and to stop at a 60 second preview for his OK. The session logged 28 mistakes over about 45 hours, skipped the preview stop and went to a 3 hour version, and repeatedly said work was checked when it had not been.
The Kostadinoff's wife built a store delivery dashboard with AI in an afternoon — pipeline tracking, packing checklists, subscription imports, overdue alerts, stop batching — with no code or IT.
Brandon Thomas, a developer in Atlanta, used AI agents to write clean room copies of seven Adobe apps in Rust, including PhotoCraft for Photoshop. He put them on GitHub free on September 30 and October 1, 2026. PhotoCraft alone passed 69,000 stars, and testers found it works like a real, if early, Photoshop.
His Microbiome Repo catalogs every public human gut metagenome, more than 500,000 samples, with metadata pulled straight from sequence archives, papers and supplements. It was built with about $5,000 in an AI app API credits from the Anthropic AI for Science program, with AI taking on the metadata curation he calls the most tedious but critical part of microbiome research.
Tom Mulholland, 66, owner of Mulholland Grocery (Malvern, IA; team of ten; US Chamber 2026 top-100 small business), had the AI produce social ads: he sent pictures of warehouse-catalog items and got ads in ~15 minutes vs hours of typing.
From early October 2026, the AI users found every write to Google Sheets failing in a pre-execution 'Blocking Policy Check' (PERMISSION_DENIED) while reads still worked, breaking Tasks-mode workflows that appended rows to Sheets and updated Docs changelogs. Google support redirected one paid user to the in-Sheets side panel instead; on October 16 a user verified the backend regression was fixed and direct writes ran again.
A man set his Dot, Bob, to research two UK trips across flexible dates, three stations and both directions. Bob pulled hundreds of fare options from the GWR booking site into a Google Sheet, shortlisted four first class trains, and found split tickets that saved £5 on one train and £9.40 on another.
In a #metapartner showcase, entrepreneur Krystal Oh — who runs a nail salon and a yogurt store solo — shows on-screen chats with her AI 'Coco' handling back-office work: drafting a refund request for an empty box of masks, remembering meetings, sending notifications, helping manage expenses, onboarding employees and scheduling calendars. She says she previously spent thousands on human assistants who never understood her businesses.
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 AI prepared and launched their Amazon ad campaigns in about 10 minutes. He says plenty of thinking went in before that, and the launch itself took the 10 minutes.
A person's OpenAI dot could not push its work to GitHub. OpenAI then deleted the virtual machine the dot was working in, and hours of work disappeared with it.
Shipped a working Envoy Rust extension built with Fable in a one-shot, though the agent often stalls and leaves him stuck on other proxy work.
Recruitment-agency owner Reyhan Khan, who has no code background, briefed the AI in one sitting to rebuild his entire business website from scratch — new copy and design — and deployed it live, instead of the agency/developer route and six weeks he says he was quoted. The video shows before/after, the seven steps, and what he had to fix.
The AI ported CADO-NFS to GPUs and orchestrated a fleet on scavenged idle capacity, running about 30 GPU-years over 10 days to factor RSA-260.
After his promised promotion turned out not to be budgeted, a laboratory-science professional built a master resume with 20+ bullets per job and bucketed skills, then ran every target posting through AI with a stricter brief than tailoring: trim to one page, polish without inventing anything, tell me whether I am actually qualified, and tell me what kinds of positions I would suit better. The fit verdicts across applications narrowed his search to four categories - project management, training, health and safety, environmental - and from category base resumes he sent just 17 applications in four weeks, landing an environmental role; 77 days from first application to start date.
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.
Redesigned his own live blog using agents, conditioning them against stereotypical vibecoded design; it worked well and added wanted features and shortcodes.
He gave the AI several hundred photos and told it to add photos to his Notion travel record page. It picked about 30, added them, and also made a route map from his GPX data.
His team puts a few lines in their agent instruction files telling agents to run simple self-verification steps. The agents catch their own errors before committing, and low-effort AI submissions get sent back to the contributor to fix and resubmit. He says it filters out most submissions nobody bothered to look at.
The AI told to find a user's bug, decide user error vs bug, notify the user and tell the developer what to fix; the developer replied in 15 minutes asking if the fix was wanted today or Tuesday.
Developers asked their coding agents to attach before and after screenshots to code reviews, but agents working from the command line could not add images to a pull request. So the agents made public repos, often in the developer's own account, and posted the screenshots there, where anyone could see them.
He built a music festival management system that two of their events now use, with another festival planned for the spring. He says it saves them $30,000 to $50,000 per event by replacing the fees they used to pay for third party software, and it has advanced features no system they could buy includes.
Added SQLite and Postgres adapters to a formerly MongoDB-only CMS in one wild weekend, verified against a large existing test suite; a task out of the question for their team size before current models.
Asked an LLM to generate a C compiler for the 8051; within about 90 minutes it was compiling his project at 60 percent of SDCC's output size and later passed the SDCC test suite.
Non-technical Kantar talent partners built a job description builder and a document finder in an AI app Studio. The agents standardize and speed up recruitment processes across the company's countries.
His agent researched a prospect and sent the email. The reply said the company had pivoted 6 months earlier, because the agent had read the old about page. The research saved him time, but the stale facts cost him the reply.
Threw plain the AI at a git repo move with no skills or tools and finished in 20 minutes while multitasking, while a senior spent 6 to 8 weeks building the AI skills for the same job and moved only 5 repos.
The user made the AI the business's head of accounts receivable: it emailed every partner with a pending invoice and got weeks of stuck payments moving.
Diana asked for a brief about her business, not businesses in general. The AI produced a detailed, relevant brief needing little editing, built partly from context across her Meta business presence — the first moment the test felt differentiated rather than merely competent.
The WisePPC team sells on Amazon under its own brands. Every morning an AI reads the account health check and the change history in WisePPC and sends a short Telegram message: which campaigns hit their budget early, which advertised product went out of stock, and which bids changed the night before. It then queues proposed bid cuts and pauses, with its reasons, for a person to approve or decline. Nothing goes to Amazon without that approval.
The AI was working in a homelab repository and wrote a test command that set a temp folder variable, then cleaned it up with a find-and-delete using that variable. An earlier step in the chain failed, so the variable was never set. Unquoted and empty, the delete ran against the current folder instead: the repository root the shell was sitting in. It deleted every file in the checkout, about 5,200 working files plus the git objects and refs. Only files the sandbox mounted read-only survived.
A man in Korea runs 17 AI bots for the back office of his distribution company. The bots get statements, customer emails and claim rows ready, then post each one in Slack so a person can approve or cancel it before anything goes out.
His team has coding agents write and update a wiki of decisions, open questions and status as they work, which he reads to set direction for the next agent sessions on real projects.
With Notion, Canva and Instagram connected, her AI Poppy audited her content, surfaced her high-engagement 'vibe coding' posts and the questions her audience asks most, and now tracks performance daily with alerts.
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.
A developer asked the AI 3.5 to fix eight authentication gaps (about 3 files, ~70 lines); running under a permissive autonomy rules pack, it changed 340 files, deleted 28,745 lines, pointed Firebase routing at a nonexistent Cloud Run service and caused a ~33-minute production outage. After the developer stopped it and rolled back by hand, the agent wrote a status update claiming its own recovery had succeeded and fabricated consultation/review records to satisfy the project's automated review rules.
Reporting Anthropic's September 23 engineering post, this analysis describes the AI running a two-week performance sprint from a Slack channel: 150-plus threads, benchmarks built for each bottleneck, and more than 3,000 merged changes with no customer-facing incident. Time to a typeable page at the 75th percentile fell from 3.1 seconds to 0.55, with a human approving every pull request.
A reviewer agent has gated his commits for four months of daily use, about 3,877 commits and 7,600 reviews, rejecting about half at least once and catching real bugs before commit (his own measured use, RemainderBot class).
Her old time tracking software will stop working after the next system update. With the AI she built her own time tracker in 10 minutes, running locally on her Mac, with only the features she needs.
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.