OpenAI and Anthropic Study What People Use Their Chatbots For

Bust out the metrics // Illustration by Kate Walker

Both OpenAI and Anthropic released user studies this week showing how people actually use their chatbots.

Why It Hits: With so many chatbots available, it can be overwhelming to know which tool is best for which task. Luckily, plenty of users have already figured out what works for them — giving newcomers a helpful guide on where to begin.

Behind The Code: According to Insider, OpenAI’s ChatGPT is primarily used for “augmentation,” while Anthropic’s Claude leans towards “automation.”

  • OpenAI found students, marketers, and knowledge workers leverage ChatGPT for writing, editing, summarization, and brainstorming.

  • OpenAI splits activity into three buckets: Doing, Asking, and Expressing. Most conversations fall into Asking, which “supports decision-making but does not produce output directly.”

  • Meanwhile, Anthropic found Claude excels in software development and enterprise automation tasks like coding, processing documents, and generating reports.

  • But this is a relatively new phenomenon — it’s the first of Anthropic’s reports “where automation usage exceeded augmentation.”

The Future: These reports will likely serve as a North Star for safety researchers, providing a starting point to identify blind spots in each bot’s specific use cases. For instance, if everyone is using Claude for coding, many companies could end up experiencing the same bugs. Or if everyone turns to ChatGPT to write their college-admission essays, universities will know the tell-tale signs to look out for.

Prediction: The split in usage suggests a future where individuals and companies rely on different chatbots for different purposes — reducing the possibility of any one tool dominating the market.

Together with 1440 Media

Today’s email was written by David Vendrell.
Edited by Nick Comney. Polled and Copy-edited by Kait Cunniff.
Published by Darline Salazar.

Reply

or to participate