The Defensibility of Large Language Model APIs

The biggest knock that I've heard right about and have chatted about with other investors is a lot of these horizontal solutions are wrappers on top of one of these large language model APIs. How do you think about modes over time? Or do you feel that it might just be discounting the execution of a team and what it takes to actually hit distribution? Yeah, it's surely the right question to ask. I think it's very much case by case. There are certain applications that you can build on top of large language models by API that truly at the local level ingest and transform and improve the data and work with it in a really interesting way. Companies like that are really interesting to