3 ms·

Claude is absurdly better at coding tasks than OpenAI. Like it's not even close. Particularly when it comes to hallucinations. Prompt for prompt, I see Claude b

Claude is absurdly better at coding tasks than OpenAI. Like it's not even close. Particularly when it comes to hallucinations. Prompt for prompt, I see Claude being rock solid and returning fully executable code, with all the correct imports, while OpenAI struggles to even complete the task and will make up nonexistent libraries/APIs out of whole cloth.

codingwagie 2y ago
Yeah, sonnet is noticeably better. To the point that openai is almost unusable, too many small errors
rubslopes 2y ago
I've been using a lot of o1-mini and having a good experience with it. Yesterday I decided to try sonnet 3.5. I asked for a simple but efficient script to perform fuzzy match in strings with Python. Strangely, it didn't even mention existing fast libraries, like FuzzyWuzzy and Rapidfuzz. It went on to create everything from scratch using standard libraries. I don't know, I thought this was something basic for it to stumble on.
- ssijak 2y ago
  just ask it to use libraries you want; you cant expect it to magically read your mind, you need to guide every LLM to what are your must/nice haves
  - rubslopes 2y ago
    Hey, I just wanted to say that I kept using it and the new Sonnet 3.5 is so much better than any other competitor.
egillie 2y ago
Does anyone know _why_ it’s so much better at coding? Better architecture, better training data, better RLHF?