@Tech_girl@dannyveigatx But if its a unique form of language it hasn't been exposed to (like the language of an Amazon tribe or a novel concept that happened after it's training) it liekly won't generalize perfectly. When that happens we get AGI and good luck everyone!
@Tech_girl@dannyveigatx corpus of "correct" text examples. This is why it can hallucinate. The model is just a function and will produce an output. But if that specific sequence of words falls out of its learned distribution, it will use its generalized understand to give you the next most likely word,
@SkylineSTEM_VPA@ColoradoMesaU@SHS_Principal You're very welcome!
As much as I want to take the credit just to mess with him, it was actually my brother Ryan, an even older alum, who presented.
If you want some insight on the computer science program though then I'll be there.
@Apple quit trying to monopolize the entire market and adopt RCS or come up with another solution. Tech is supposed to be a tool to minimize boundaries not create them