An LLM cut 60,000 useless lines from his test suite and fixed bugs it found, in about 2 hours
LLM audited about 100k lines of tests, deleted about 60k useless lines, wrote broader scenario tests, and caught and fixed bugs, in about 2 hours.
A random mix from the whole library. Shuffle for a new set.
LLM audited about 100k lines of tests, deleted about 60k useless lines, wrote broader scenario tests, and caught and fixed bugs, in about 2 hours.
Daren runs his film company alone. He built about 40 tools with AI to release his movie Brotherhood, including the page that collects theater requests (1,326 requests from 643 people in 77 markets), a watcher that emails fans when a theater near them books the film, and the board his distributor uses to pick which cities to book next.
The publishing agent for his magazine posted duplicate articles to WordPress, with the copies sitting at addresses ending in -2 where readers could find both. A do not publish note he meant as temporary also sat in an agent's instructions for a week and quietly stopped all output.
His AI pulls from GitHub, Jira, Asana, his calendar and his email, writes up his week, and posts it to Slack without him asking.
A new feature for his blog, built using his voice with an agent, now in use.
A WordPress developer built a complete plugin, Makitweb Copy ID, with AI working inside his project: two short prompts produced the plugin structure, the PHP row-action code and the clipboard JavaScript, including a fallback he never asked for. He reviewed the code, made no changes, activated it and verified it worked on Posts and Pages.
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.
Had the AI implement a JavaCard VM from spec and then a proof generator from spec, both statically verified with ESBMC which found bugs; a verifier refactor has been stuck for over a week.
Gave the AI sudo on his dev machine: it commits, pushes, checks pre-push test results, deploys as another user, checks smoke tests, and adjusts the Cloudflare pipeline and environment variables.
At the start of a 2 hour 7 minute TGV ride from Paris to Valence, his shop did not exist. Talking with AI, connected through a new MCP link and without opening the WiziShop admin once, the shop was online by the time he arrived.
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.