A random mix from the whole library. Shuffle for a new set.
He set up an AI executive assistant that keeps his appointments and reminds him two days before, one day before and on the day, with push notifications to his phone. It also tracks his house todo list and his government social benefits paperwork, and turns things he just says in plain English into structured records.
The AI or the AI built an embeddings plus logistic classifier for his emails reaching 95 percent accuracy from 50 to 100 examples, trained in under 5 minutes locally.
In Diana's inbox-prioritization test, the AI produced a useful organized view but concluded she hadn't sent an email she had in fact sent — the thread structure and CCs confused it. If she'd trusted the first answer, she'd have sent a duplicate email.
In two weeks of real inbox-and-calendar work, The AI Crux had the AI book three conference rooms. Calendar work was functional but conservative: it refused anything outside 9-to-5 local without an explicit override, which the reviewer framed as a guardrail more than intelligence.
A Tesla engineer got an email that passed SPF, DKIM and DMARC and showed a verified check, but felt off, so he had his AI review it without clicking anything; the bot found a hidden phishing link, explained why the standard checks had passed, and wrote a report to xAI with a proposed fix. He then had it add a standing routine that flags likely phishing in his inbox, which has since flagged more messages.
A builder used AI to make a 'describe your situation' tool over Australia's official welfare pages: it takes a one-sentence description, works in 24 languages and shows which Services Australia payments might apply, each linked to its official page. Across 43 test questions it invented zero payments and refused correctly 100% of the time, with payment recall of 0.76.
He connected his Gmail to an AI task that watched every new email and was only allowed to interrupt him when a message needed action or came from family, with a draft reply ready. All afternoon it alerted him three times, for a bank message, a family email he answered from its draft, and a ticket itinerary change he needed to act on.
A man asked the AI to prepare a compliance letter to a US state agency to get his driver license reinstated, and to send it from a named email account. He had typed his license number one digit short in his note, the correct number was in a photo in the same folder, and the AI itself had flagged the mismatch. The letter still went out with the unverified number in the PDF and the email. The correction was then sent from a different account than the one he had named.
An operator had the AI set up a set of websites that pull publicly available but painful-to-access government data, present it in easy-to-read form, and send weekly emails - turning civic data nobody uses into a small publishing operation.
Le Monde's testers (Oct 20, 2026) connected a rarely checked secondary Gmail account and asked Instinct to scan twelve months of mail for anything needing attention. It flagged exactly one item, a useful one: a subscription about to expire because the bank card on file had expired and the renewal alerts had been missed. In the same test Instinct compared door-manufacturing services, found an installer, sourced freelance journalist profiles on LinkedIn and found a cheaper train ticket; the testers stopped short of connecting their sensitive accounts.
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.