Re:How Good Is ChatGPT at Coding, Really?
by nutshell89 on Saturday July 06, 2024 @09:56AM
(#40890502)
The study involves AI based code generation on Leetcode problems using GPT-3.5 — functional code was produced between 0.66 percent and 89 percent of the time, but I'd like to see the same study conducted with GPT-4 or the latest Claude models.
[ Parent
| on HN ]
There are no comments on this comment yet.