The policy summarizes itself this way: It’s fine to use LLMs to answer questions, analyze, distill, refine, check, suggest, review. But not to create.
Saying that “it’s fine” seems a little overly broad, because sometimes it isn’t, but I suppose that by “fine” they simply mean that it’s not forbidden by the policy. A more complete summary would also mention that where it is allowed, disclosure of LLM use is required.
Personally, on the few occasions where I’ve resorted to querying an LLM it’s been very much the same kind of experience that using Stack Overflow used to be. It’s a quick way to see one way that other people have done some thing, but you wouldn’t want to rely on it being the right way.













As was said elsewhere on the Internet: