Introduction to LLMs
Large Language Models (LLMs) have been gaining popularity, and a recent study by OpenAI has shed some light on how people are using these models. The study reveals that LLMs are not replacing search engines, but they are changing the way people access and consume information.
How LLMs Work
A chatbot is a statistical model trained to generate a text response given some text input. The more advanced chatbots have a two or more-stage training process. In the first stage, LLMs are trained to predict the next word in a string. This stage is crucial in helping the model understand the language. The second stage is where things get a little fancier, and models are trained to generate "quality" responses to a prompt.
Key Findings from the Study
The study found that the top three use cases for LLMs are Practical Guidance, Seeking Information, and Writing, which account for 80% of all conversations. Non-work-related usage is increasing, with 70% of all usage being non-work-related by July 2025. Writing is the most common workplace application, accounting for 40% of work-related messages on average in June 2025.
Use Cases for LLMs
The top three use cases for LLMs are:
- Practical Guidance: This includes asking for advice or guidance on a particular topic.
- Seeking Information: This includes searching for information on a specific topic or asking questions.
- Writing: This includes using LLMs to write or edit content, such as emails or articles.
Non-Work-Related Usage
Non-work-related messages grew from 53% of all usage to more than 70% by July 2025. This suggests that people are using LLMs for personal purposes, such as seeking information or entertainment.
Workplace Applications
Writing is the most common work use case, accounting for 40% of work-related messages on average in June 2025. About two-thirds of all Writing messages are requests to modify existing user text rather than create new text from scratch.
Impact on Publishers
The study’s findings have significant implications for publishers. With LLMs changing the way people access and consume information, publishers need to adapt to these changes. One of the most practical outcomes is the apparent change in intents. For eons, publishers have been focused on navigational, informational, commercial, and transactional intents. Now, they need to consider "Doing" or "Generating" intents, which are huge.
Takeaways for Publishers
Publishers need to build linkable assets that add value, such as tools and content that can’t be synthesized by machines. They need to focus on creating high-quality content that adds value to users. Programmatic SEO can drive amazing value, as can tools that answer users’ "Doing" queries time after time.
Conclusion
In conclusion, the study by OpenAI provides valuable insights into how people are using LLMs. The findings suggest that LLMs are changing the way people access and consume information, and publishers need to adapt to these changes. By focusing on creating high-quality content and building linkable assets that add value, publishers can thrive in a world where LLMs are becoming increasingly popular.