Read in Other Languages
A Driver, an AI, One Night: How a Recording System Was Built (Summary)
On the night of September 26, 2026, in Toronto, a rideshare and delivery driver sat down with an AI and asked one angry question: why are the consoles in my OBS scenes empty, and did you break them? The answer was in a single line of the OBS log: two OBS processes had overlapped for three seconds and the browser engine never started. Nobody was at fault; timing was.
Two and a half hours later, at 00:42, a complete recording system existed: a Day Log on the phone (odometer, fuel, charging, breaks, notes, rides), a one-button DRIVE flow that checks OBS, measures its health, prepares the scenes, starts the recording and plays the intro, a STOP flow that reminds the end odometer and plays the outro, app buttons for Uber, Lyft, DoorDash and Uber Eats that change the on-screen badge, every event pushed to the right console and the subtitle bar in seven languages, Siri shortcuts, a sensor feed that rides on the first human tap, and a rules file written so that another AI session can read it from scratch.
Along the way the AI gave one wrong answer ("AirPlay only works on the local network"), was corrected by the human, verified it on the machine and admitted the mistake. The human made every decision: recording instead of live streaming, masking afterwards, no customer name or address ever on screen, OBS closed properly and never killed, an interface entirely in English.
The human side is the point. In Muaz's own words: "People make up imaginary AI stories on YouTube and earn money; I live with AI, and I'm crawling." There is nothing imaginary in this one: a real night, a real recording, a real car. The AI is the tool; the human is at the wheel. The full text of the story is on muazturkyilmaz.com; the video is on the Fusion Analysis channel.
Full story: https://www.muazturkyilmaz.com/ · Channel: Fusion Analysis (YouTube)