What do people use Claude for
So far, I had seen only high-level statistics or anecdotal evidence on what people actually use their LLM for.
Anthropic has now published a study addressing exactly this question, providing more detailed insights based on an anonymized database they built. While the results are interesting, they remain fairly aggregated, and honestly, contain few surprises. 😌
The key use cases are (my selection):
Programming - no surprise here, as this is already a well-established use case being piloted by numerous companies
Content creation & communication: again, as expected. LLMs excel at crafting marketing copy, improving style, and finding the right tonality and structure for specific target groups
Academic research & writing: I would have preferred more specificity in this category. Are we talking about content creation for academia, or actual research, including question definition and research design?
Education and career development: here too, more detail would be valuable. The category feels overly broad - does it encompass content help, exam preparation, curriculum development, or all of these?
Business strategy and operations: This is particularly intriguing, but I'd love to explore another level deeper. Unfortunately, the research tool Anthropic developed isn't publicly accessible
Interestingly, the authors highlight significant variations in use cases across languages. While coding maintains consistent popularity across regions (representing 15-25% of conversations), there are distinct regional patterns. Japanese and Chinese users show notably higher engagement with topics around aging populations and elderly care, while Spanish-language users demonstrate particular interest in economic analysis and rural development. This suggests that local contexts, concerns, and interests significantly shape LLM use - there isn't a single, homogeneous AI use case that fits all users.
In summary, we appear to be leveraging Claude effectively, with primary use cases clustering around programming, writing, and research - areas where LLMs can provide substantial assistance. Beyond these common applications, there's a fascinating long tail of diverse use cases spanning creative writing, gaming, and more. Anthropic shares several surprising, fun, and decidedly nerdy examples that illustrate just how varied LLM usage can be:
Dream interpretation
Analysis of soccer matches
Disaster preparedness
"Hints" for crossword puzzles
Dungeons & Dragons gaming
Counting the r's in the word "strawberry" (this has become something of a meme, following ChatGPT's notorious difficulty with this task)
As always, I'm interested in your thoughts: Feel free to message me. Or, if you prefer, you can share your feedback anonymously via this form.
All opinions are my own. Please be mindful of your company's rules for AI tools and use good judgment when dealing with sensitive data.