Thanks!!! Will try to find time to do some more videos!

I think the ability for OpenAI’s GPT models to ingest a long context length like 4k, 8k and now going up to 32k is something very unique and powerful about their models that I have not seen elsewhere. So while sending in so much data in the prompt will definitely incur additional computing resource, I think they probably treat it more as a feature than a bug. One must be on a paid program (like GPT plus to use plugins on the chatGPT interface) and for API use, they charge on a per token basis rather than a per API call basis. In fact, OpenAI’s announcement yesterday of the Function Call feature in their API makes it even easier to send your own data in. Also they keep pushing the boundaries of their context window length. So my guess is they want people to make full use of it and, of course, pay for it!

@@hubel-labs Oh OK, thanks for the additional information on that. I didn't realize that you were using a paid API for that, because otherwise it would probably be a matter of time before someone took advantage of that. This video was awesome and look forward to your others!