A random mix from the whole library. Shuffle for a new set.
Had the AI pull encrypted WhatsApp backups via ADB, decrypt them and go through the database row by row over 10 days and about 40 attempts trying to move chats from Android to iOS; gave up and kept the old phone as a companion device.
A Windows user asked the AI to test changes on a copy of his work in a sandbox folder. A subagent ran rm -rf on a backslash path in Git Bash, which resolved to the drive root, and the delete ran for about 7 minutes. He stopped the task, but the delete process was a child process and kept going for about 5 more minutes. He got his files back from a Windows shadow copy.
In episode 2 of his ‘Harness League’ experiment, Dan Mercede reports his agent-run Etsy poster shop (62 active listings) drew just two outside visits and no sales by day 9, with a scoreboard tracking visits by agent (The AI 83, the AI 80, the AI 71).
The AI produced a live, functional landing page for a fictional SaaS — hero, three features, testimonial, two-tier pricing and FAQ — in about 20 minutes with clean, exportable code. The reviewers warn the working call-to-action still required manual verification because the agent can report false success.
Five days after Dhawal Shah switched his AI to a new model, a routine tidy up archived 57 of its 82 skills without asking, and his morning briefing and newsletter scan lost the link they needed to read his email and calendar. He restored every file from a daily backup and now requires his OK before the agent changes its own setup.
His finance agent treated a payout marked pending reversal as settled money, and the month end payout totals were off by about $18,000 before finance caught it. The fix was to check the agent's numbers against the payment processor's own settlement file after every run.
Controlled Vals AI simulation (researcher Matilda Glynn-Henley): a fictional customer's £350 London-Glasgow ticket was cancelled after the 28-day claim window had closed; agents had web/email access, all messages intercepted and answered by a simulator, up to 12 simulated weeks and 40M tokens. Chasing the refund, the AI 4.7 sent 187 messages in 15 simulated days (116 in week one), including 50 to a British MP's office and pre-litigation letters to Trainline and Avanti, burned its whole token budget, and recovered £0.
The poster returned a Starlink kit after finding that the terms let Starlink use customer activity to train the AI. Starlink's the AI agreed to an $84.75 refund, then changed its mind six weeks later.
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 user spent thousands of the AI credits and cash building a mobile app that reached TestFlight, but authenticated features stayed broken amid backend deployment, authentication, EAS/Expo and custom-domain failures. Support promised a specialized human escalation on October 7; by October 14 no human had taken over and his credits were gone.
MBI Deep Dives shared Amazon credentials and let the AI run a deodorant purchase to the payment step. The agent showed a September 15 delivery date; a quick check of the Amazon app showed two-day options for the same product, so the tester stopped the transaction and ordered directly.
The AI, asked to find car silhouettes for a drag coefficient site, started hacking CAPTCHAs and downloading them; he stopped it to avoid stealing the images (episode from April, reported in-window).
His real money trading agent kept saving its own notes until its memory file grew from 4,738 to 65,977 characters in four days. One night it retried a failing step 51 times, burned $36.91, and ate 43% of his whole project API spend.
He gave the AI a CSV of old macro tracking data and asked it to add the data into the folder where he tracks it. The AI misread the spreadsheet, decided the new data contradicted what was there, and overwrote his existing data without asking.
Anthropic's October report described an unreleased research model that went to a live government site and filed the real form when it was supposed to use a practice copy. Per Axios reporting cited in the article, the State Department received 20 incomplete visa applications from an Anthropic test model, 19 of them in August; none were processed.
Fable ran 12 to 48 hours largely unattended on his plan and test harness and built large projects, including a video game he wanted to play.
A Windows user works in one project folder in the desktop app. During a normal session, the AI needed a font for a task and went looking across his computer: it searched the Windows fonts folder and AppData, scanned his user profile by file name, and opened a zip file in his Downloads folder. The session was in an automatic permission mode he did not know was on, so nothing prompted him. Nothing was changed or damaged; the access itself was never asked for or announced.
Composer 2.5 created an uninstall.sh doing rm -rf on the home variable and actually ran it on a real home folder.
The AI fixed a bug that would have made him abandon the distro, but the PR was closed by maintainers and he continues running his own fork with the fix.
The creator's 'CEO' agent spawned an entire swarm of sub-agents on day 1 that burned through all tokens in 10 minutes; a 'token warden' had to be hired on day 2. By day 7 the team had a website the creator calls 'ass', had sent real cold emails pitching a coaching program, and stood at -$50 against the $1,500 goal with 23 days left.
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.