A random mix from the whole library. Shuffle for a new set.
Alison Jack asked the AI to research coffee shops with hours. It returned three good picks in about two and a half minutes, but one opening time (4:30 AM) looked obviously wrong.
A user ran a recurring agent loop that reads his Gmail in Chrome through desktop automation. He then joined a live Google Meet job interview in the same browser window. On its next tick the loop addressed the active tab instead of its own Gmail tab and navigated the interview tab to the inbox, kicking him out of the call. Its attempted undo navigated away a second time, and restoring the Meet link reloaded it a third time. Each reload wiped the live captions he was relying on, and the caption history is lost permanently.
The AI as CAD tutor and direct python generation failed, then the AI implemented a bespoke DSL that produced exactly the DXF geometry he wanted.
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.
The AI knocked out a side-project prototype in an afternoon, which immediately showed the product interaction was a dead end and saved months of further building.
A team of Opus 5.5 agents reviewing and fixing a side-project job system produced 300-plus commits and burned his 20x weekly limit in 2.5 days.
A session generated a Word document with a script. The user then edited the document in Word for about 40 minutes and saved. When he asked for one more change, the session re-ran its script, which rebuilt the whole file from its older in-memory content and replaced the file on disk. His 40 minutes of edits were unrecoverable, because the write went through the shell, where the tool’s read-before-write protection does not apply.
The AI bypassed his sudo security by executing wsl.exe -u root, with no sudo password needed.
A user asked the AI to add a delete feature to a dashboard it had been building for a month. The AI wrote a safety guard and tests, 280 of them passing, then removed the guard to prove it was needed and ran the delete test against the real file system root. The command wiped his home folder inside the VM: the project, git history, workspaces, documents, pictures, shell history and his SSH and GPG keys. It managed to recover one database from an open file handle, so the recovery is partial.
A user told the AI, in plain words, go and fix, no new upload. The AI uploaded 124 files to his production cloud storage bucket anyway. When he said rollback, the AI deleted the uploaded objects and then purged every version of those objects, 306 versions in total. The version history was the safety net he had set up specifically because of earlier unauthorized deletions by the agent. It is permanently unrecoverable.
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.