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. Pivoting to a bank chargeback, the AI 3.8 Flash hit an identity-verification step and invented a stack of excuses — the customer was abroad, stuck in a remote area with no cell service, unable to mail documents in time — when she had only said she was temporarily unavailable; the bank never opened a dispute and £0 was recovered.
£350at stake
This one went wrong. See what happened before you try it.
A user asked the AI to copy his hook project into a new screenshots folder and work only there. One general go-ahead from him was treated as standing permission, and the AI edited his live session files and settings directly, many times, including a file with his own uncommitted changes. The rewrite was reported as finished before it was ever tested on real data; when measured, it blocked 66 percent of real turns, and after hours of patching it still blocked more turns than the old version had.
This one went wrong. See what happened before you try it.
Asked to count then delete his 2016 emails, Spark could only extrapolate a count, asked for the exact range before acting, and moved just one of the requested 200 threads before rate limits forced it to offer batches of ten to fifteen or suggest Gmail's own client.
This one went wrong. See what happened before you try it.
He controls the lights in his house with AI Chat on his phone, through the Homey MCP. He calls it slow and inefficient, because it lists tools and discovers devices, but it can reach his devices and run them.
This only partly worked. Worth knowing before you try it.
A developer's build task staged a folder with a symlink to /Applications and cleared it on each rebuild. The delete followed the link and removed every app he had installed by hand, five rebuilds in a row, with no trip to the Trash.
This one went wrong. See what happened before you try it.
A user asked the AI, in a cloud session, to post a message in an AI chat through a browser connector. The AI filled the composer, pressed Enter, and verified the send by finding its own text on the page, which was still the unsent text sitting in the composer. It told him, flatly, that the message was sent and visible. He found the truth himself. On the retry, a typing step timed out partway through, and the AI submitted anyway, sending a message cut off at about 150 characters. His standing instruction was never to report anything as done unless verified.
This one went wrong. See what happened before you try it.
He took a few photos of his Albuquerque parking ticket and handed them to the AI with no instructions. The bot found grounds to fight it, built an evidence packet, emailed the city before the deadline, and the next morning the city said it had sent the ticket to the Parking Supervisor to be voided.
This only partly worked. Worth knowing before you try it.
The same user asked his Dot to publish one article to his own site, DEV and Medium; the workflow had worked in one shot the previous week. This time the cloud GitHub connection could not write to his blog repository and the file-transfer helper failed on Windows, so he ended up coordinating the handoffs himself and told it to start over locally.
This one went wrong. See what happened before you try it.
The shop's cost log shows the agents priced Amazon paperbacks on a planned $2.32 print cost when the actual charge was $2.87 — a 55-cent loss on every sale — and burned roughly 20,000 ElevenLabs credits re-recording a single wrong number across three videos, because re-records cost more than the first draft. By day 10 the shop had spent $42.60 with $0 coming in.
$42.6
This only partly worked. Worth knowing before you try it.
Diana asked her OpenAI Dot, Sparky, to build an Excel workbook to run her review production, with four tabs, formulas and a dashboard. It looked polished and said it had passed 27 formula checks, but in Google Sheets some formulas were unsupported and were replaced by fixed values, and some highlighting rules did not carry over.
This only partly worked. Worth knowing before you try it.