PSA: Always specify the LLM model

There are many comments here on HN (and elsewhere) where people talk about "ChatGPT" or "GPT" without specifying whether it's GPT-3.5 or GPT-4 (or others even). Then other people point that out and say that if they haven't tried GPT-4 yet, they should.

And there are likely many more where it's not explicit, but a discussion is still taking place between someone who's only used 3.5 and someone who's used 4, and they're disagreeing about capabilities.

And there are other LLM models out there, so "LLMs" is non-specific, which can sometimes cause imprecision or confusion in a discussion.

So it would be great if people who haven't tried GPT-4 yet and are willing and able to do so, did so. And if you haven't, don't want to, or can't, then provide that clarification in the discussion.

In addition, it would help if everyone stopped only saying "ChatGPT" and started saying "GPT-3.5" and "GPT-4". (And yes, there's 3.5 vs. 3.5-turbo, but I don't want to ask for too much.) ChatGPT is mostly a frontend to the models. Of course if that's what you're talking about, or about the customization OpenAI might have made to the models used by ChatGPT, that's fine.

Let's elevate the discussion around this topic together!