I noticed in rahul's video about debugging with the iOS example there was a snippet around 80% through the vid where Alex used chatGPT 4o to help with debugging
I've found o3 to be 100x better at debugging than 4o and im curious to see if o3 would help here. It's ability to reason and search the web is really helpful!
My initial guess is probably not much bc LLMs are not good at swift/ios dev but still curious to see
Oh dang, I should be using o3 to debug then - Thanks for the tip! I was using 4o as that's the default 😅.
It is also true that LLMs aren't the best at Swift/iOS development as the training data pool isn't huge, especially for SwiftUI.
This is good to know. It's always been on my todo list to better understand the nuances of each model and what tasks they perform better at. I can never keep up!