I raised some questions regarding the use of the LLM (AI) during the weekly analyst calls. Turns out after reading a bit more, I realized that the LLM is used for the potential investors to query questions about a project or token, for example, what's the total supply, what's the vesting schedule, etc. So, the LLM has nothing to do with trying to predict price of the token, it's not used to process time serie data, it's used to process whitepaper, website, and eventually social sentiments which are all text based.
I think that's a good use of AI technology that should be commercially viable. Hopefully they can optimize this LLM with a graph neural network in order to compare two different projects in the same category in a way that make sense. LLM can be very dumb, which is why I think a graph neural network may eventually upgrade the intelligence of LLM significantly. Meaning it could give meaning to words and their context, not just offer a word probability matrix.
Anyway, this project is not strictly about the AI, but I just like to make sure projects don't just slap a trending keyword but based on flush. In this case, I would say it's a good use case of AI that is achievable and could be significantly improved over time. Show Less
4
