Overheard a dev say GPT-4o is 'basically useless for debugging' and it got me thinking

I was at a coffee shop last Thursday and two guys were arguing that Claude is way better at finding bugs than GPT-4o. Made me wonder if I've been wasting time with the wrong tool. Has anyone else noticed a big difference between them for fixing code?

3 comments

3 Comments

wilson.sam2mo ago

Honestly, GPT-4o has saved me hours finding dumb typos and logic errors in code.

grayc272mo ago

Man, that hits close to home - been there myself.

grace_allen2mo ago

Honestly used to be skeptical but it won me over.