Quote:
Originally Posted by natkejs
Been having a lot of fun with AI, but definitely not on the level as some would suggest. Suppose I'm not a master prompter yet.
Most of the time it seems to mess up one thing or the other, but it's been great for getting an outline and then you'll have to go through it again and point out all the errors. While it's not maybe fully living up to the hype (at least in my experience), it does save a lot of time and you don't need to write silly code anymore. Plus it's nice to have an assistant that is lightning fast and never becomes tired of making small changes etc.
But anytime I tried to refactor more complex code or whatever it completely fucked it up, not even close to making it work. Just a step or two maybe in the right direction  |
I can't comment on what it did a few months back but this new 4o version is super nice for what I'm doing. If it fucks up I just post the errors back to it and it fixes them no problem. If the fix doesn't work it adds some debugging info, asks for the response and fixes it then. It also has memory and will pre-empt what you are trying to do, it's getting to the surreal point.
So far I have only worked in Python to recreate some scripts I'm tired of paying for updates on but it has done them perfectly (better actually) and is able to data mine from other websites / provide code to data mine from those specific sites.