From asking to doing: How the world is putting ChatGPT to work

OpenAI has a new dataset it's calling Signals, and the headline finding is less about clever prompts and more about a shift in behavior. People aren't just asking ChatGPT questions anymore. They're handing it actual tasks—drafting documents, organizing plans, working through problems that used to require opening five other apps first.

The country-level breakdown is where things get interesting. Adoption isn't uniform, and it never was going to be. Some markets lean heavily into ChatGPT as a work tool, woven into daily job tasks, while others still treat it mostly as a curiosity or a homework helper. OpenAI frames this as evidence of a maturing product, one that's moved past the novelty phase in places where usage has had more time to settle.

What's notable is the direction of travel, not just the current snapshot. Early chatbot use everywhere tends to look the same: trivia, quick facts, maybe some creative writing. But as populations get more comfortable, the pattern OpenAI describes is consistent—queries turn into workflows. That's a meaningfully different kind of product usage, and it says more about where AI adoption is headed than any single benchmark score does.

OpenAI didn't publish exact figures for every region, and it's presenting this more as a trend report than a full transparency dump. Still, the company is clearly trying to make the case that ChatGPT has crossed some threshold—from a search-engine substitute to something closer to an assistant that actually executes. Whether that holds up under scrutiny elsewhere is a separate question, but the framing itself tells you what OpenAI wants the narrative to be going into next year.