Life admin

Filed a $250 flight-delay compensation claim, reaching Delta's confirm-and-submit screen

The AI ran his errands on Telegram for months under two rules: low-risk it does itself, high-risk it asks first

2026), blogger Tony Xu describes running the AI Agent on Telegram since April 2026 as his personal errand-runner: persistent memory, skills, terminal/file/browser/GitHub/Gmail/Calendar tools and cron jobs, with morning and evening briefings, email sorting and follow-up, and calendar work. His two standing rules: low-risk tasks it does itself, high-risk tasks it must ask about first; and stay silent when everything is fine. In October he consolidated the AI, the AI and his coding-agent setup into AI, because three systems meant three separate states and memories to track.

Had the AI auto-fill a Sarvam DevOps Engineer job application: read the form

Hemant Kadam, demonstrating his 'craziest use case' after a couple of days on the AI, opened the app to a Sarvam DevOps Engineer application page. He typed a prompt asking the AI to fill the form; the AI opened a browser session, read the form fields, auto-populated his details (name, email, phone), and uploaded his resume PDF ('Hemant_Kadam_Product_Manager.pdf').

My agent cancelled my Mint Mobile account over chat when they said I had to call

She asked the AI if it could cancel her Mint Mobile account. The company's chat agent had told her she needed to call in. She gave the AI her Mint details, the AI opened its own chat with the company and got the cancellation done, with no phone call and no talking to people herself.

Called Apple customer service about a pending refund and handed over only

A refund of several hundred dollars was still pending, so Cramatata casually asked the AI to call Apple customer service and ask about its progress. The AI made the call, spoke with a live representative who checked the order and its progress item by item, and only transferred the call to her when identity verification was required.

Publish the government data nobody can read, and email it weekly

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.

I used AI to build and file my own court complaint

She says she used AI, cross-checked by two more AIs, to put together a legal complaint and file it herself, without a lawyer.

The AI filed the expense claim for his AI subscription

Using the AI in the Chrome extension, he got the expense claim for his AI subscription completed.

My bot drafts Community Notes on election rumors, and I approve each one

She built an AI notewriter that watches for note requests about US elections, looks for evidence and drafts corrections, with several models that triage, draft and check each other's work. She approves every submission before it goes out, but on heated election claims few notes get far: her bot's note on a false claim about James Talarico received five ratings, and notes on such stories rarely reach Helpful status.

The AI built a label and filter system for a crowded inbox

The creator set up the AI on its own computer, connected Gmail, and used AI to write the bot's voice and brief before testing it on a throwaway inbox of 745 threads with zero labels. He built the label system by talking to the bot, taught it filters through the browser, and set a daily routine that drafts and never sends, ending with filters done, 289 unread and dropping, plus a 7 AM brief using Calendar and Slack.