>>1003571
to a very basic level, ie reading the manual but using chat gpt to read it for you but hallucinate on top of it. You have no idea what you are talking about basically

It seems like the next logical step here would be that it gets integrated into the programs directly, and then instead of asking it how to do the thing you just tell it to do the thing. But with how unreliable these things are it would surely do the thing wrong, perhaps so wrong that it causes a crash- and then you won't be able to fix it because you're using the AI as opposed to actually learning and you don't actually know what the problem is. But in any case some day it'll reach a zenith of perfection and you'll be able to effortessly prompt entire movies by directing a different AI through each step of the process. I'd say about ~2 more weeks

>>1003571
Great. I'll look at it tonight and tell you if it's got the correct methodology

>>1003571
How good is it at niche programs though?? Having an AI to help you with blender or any other software that already has 1 million tutorials online is pretty much worthless, just a little bit faster than googling for it. I can only imagine something like this being genuinely useful for learning the ins and outs of programs that you can find less than 10 videos about on youtube. Although I wonder how it would be able to instruct you on that since it barely has any data to train off of

>>1003625
>Although I wonder how it would be able to instruct you on that since it barely has any data to train off of
It wouldn't be able to do fucking shit. AI is just the latest scam. The amounts of hallucinations that occur is completely out of control. AI is the absolute worst