Well video doesnt get fed in to #leoai, at least as of now. That is likely too expensive and really only the LLMs are multi-modal. But the more text that we feed it, the better.
Different topics, languages, and obscure facts can really help.
You are viewing a single comment's thread from:
Okay, so at the moment, everything that is text we should feed. Okay, got it.
Yeah. LeoAi doesnt have the resources to go through video or images. That takes a lot. Right now, each thread that is added automatically updates the database.
Good to know that. It is to continue with our work and feeding the database.
Yes. The ability to train this is crucial. We have to understand that it is all being fed in. People are going to be upset when they do not get the results they want but it is mostly their fault...need to keep pushing data in.
I understand, I will do my part.