It wants to seem smart, so it gives lengthy replies even when it doesn’t know what it’s talking about.
In attempt to be liked, it agrees with most everything you say, even if it just contradicted your opinion
When it doesn’t know something, it makes shit up and presents it as fact instead of admitting to not knowing something
It pulls opinions out of its nonexistent ass about the depths and meaning of a work of fiction based on info it clearly didn’t know until you told it
It often forgets what you just said and spouts bullshit you already told it was wrong
They probably did anyway.
It does not matter. Output is based on probability.
Thanks, way to forget that it’s fancy autocomplete.
Yeah they probably pirated it, but apparently done weight knowledge sources very well.
That seems to be the big missing part of all this gen AI.
I wonder if the selection of images online tends to be the higher quality subset of all imagery, whereas writings are all over the place, including a large quantity of shitposts. Could it make training image generators easier than text ones?