Of course the percentage of people who have ever used a chatbot is going to go up over time.

Yes, and crucially, it never goes down.

It's like asking: "do you use a motor vehicle?" And then counting everyone who has ever been in a car, a truck, a bus, or potentially even a train as a yes. It plainly conflates active users with exploratory or incidental users.

The only reason to do such things is to inflate numbers because being honest about them makes it look bad.

Yup. I used Gemini like a year before it took off back when Google first sent out testing invites, because I wanted to see what it could do. That's very different from active use, but this survey would count me as a regular user.

Although since Google shows me an AI summary whenever I search now, maybe I count as an active one anyways... along with everyone else who uses Google.