https://arxiv.org/abs/2308.02312
According to this study from Purdue that compared ChatGPT with Stack Overflow: "ChatGPT’s answers are more incorrect, significantly lengthy, and not consistent with human answers half of the time." However, the only location where I see the version mentioned, it is 3.5: "For the additional 2000 SO questions, ChatGPT 3.5 Turbo API is used." GPT-4 is much better than GPT-3.5 in programming, so this study might already be outdated.