0 comments
OpenAI’s latest usage report on ChatGPT sheds light on how people are moving away from having the chatbot complete ‘doing’ tasks, but are leaning on it for advice, problem solving, and outsourcing judgement calls. Is this cause for concern?
You know who doesn’t ever get brain fog? ChatGPT.
Use of OpenAI’s chatbot has become nigh-on ubiquitous, and people are leaning on the technology for just about everything in 2025. It’s gone from a barely entertaining gimmick to a life fixture for countless people within the last year.
The company’s latest analysis, based on 1.5 million conversations between July 2024 and July 2025, shows how habits with the chatbot have evolved.
Writing tasks, once dominant, dropped from 36 percent of usage to just 24 percent. By contrast, information-seeking queries nearly doubled from 14 percent to 24 percent, while requests for practical guidance — tutoring, troubleshooting, step-by-step instructions — remained steady at 29 percent.
Arguably the most pertinent stat, advice-driven interactions, now make up more than half of all ChatGPT usage. In OpenAI’s terms, ‘Asking’ has overtaken ‘Doing’, essentially marking its pivot from ghost-writer or PA to life-coach. It’s what the people want.
This change is easy to understand. ChatGPT is quick, confident, and endlessly available. It offers ready-made clarity where the internet often provides noise. The problem is that clarity is not the same thing as critical thinking. Where even Google requires some effort — sifting sources, cross-checking, identifying bias — ChatGPT delivers conclusions neatly packaged in a single paragraph, usually rounded off by a nice kiss-arse sentiment.
Ironically, you could argue that convenience is the product, while also being the major trade-off… like a lot of things in 2025.
