A comment should be evaluated on its own, not based on who it was written by.
I try to avoid looking at the names until after the discussion is over. When I remember the fact I'm posting among giants, I find it difficult to post at all. HN is a place where the person who created an algorithm shows up to explain it to you. The fact such people post here is the number one reason to come here.
For example, if a comment claims an article is wrong about the development and use of the Nagle algorithm, I am very much not well-equipped to evaluate the comment on its own since that is not an area I know much about. If I see that comment was written by Animats, though, I can be reasonably confident I should trust the comment over the article, since he invented the algorithm.
If someone makes the statement that something in an article is incorrect, and fails to takes into account some minutiae, it is probably because the person is knowledgeable about the topic and believes it to be the case.
There is no need to research who the author is, or to mod the person down because he appears to be a contrarian that doesn't follow the hivemind. If anything, that's just an interesting subject to look into.
It is, but intent is distinct from correctness. I can believe that everyone is commenting in good faith and stating their honest beliefs, but that is little help when commenters take contrary positions and I would like to determine who is correct despite not being knowledgeable myself.
Under those circumstances, authorship is a useful signal. It doesn't need to be the sole determining factor, of course, but it's better than nothing.
I think it best to make your own opinion based on the arguments presented to you. Sure, someone may convincingly say wrong things, and you might end up believing them, but then it's up to you to develop your own ability to see through bullshit (of which there is a lot in the tech space, especially when you get close to VCs).
I'd hope most people would agree with you, since fame is not the same as expertise. A well-known author who happens to talk about a subject they are knowledgeable about can be worth paying attention to, but the same author talking about a completely different field may (should?) warrant more skepticism. Authorship can be a useful signal, but that doesn't mean it is always a useful signal.
> but then it's up to you to develop your own ability to see through bullshit
I mean, yes, that is ideal, but unfortunately but not all BS is equally detectable. Sometimes it really does take intimate knowledge to know someone is incorrect, and achieving that for everything just is not realistic for most, if not all, people.
Initially I got the most downvotes I think I'd ever seen†, and then Thomas realised oops, yeah, he's not saying what I thought he's saying and it shot back up over the course of an hour or so. IIRC the auto-transcripts for example transcribe PKIX (an RFC explaining how we shall use the X509 certificate system for the Internet, and the regime resulting from that RFC) as "Peacocks". Hilarious. But, not a criticism of the person on the show (Ryan Sleevi I believe, who I rather like although we've never met in person).
I have written things I don't feel good about, especially trolling groups who believe something I don't, and deserved to be downvoted even if they were strictly true. But those mostly get -1 or -2 or something. Thomas' fans just keep hitting that down arrow apparently.
Other than disagreeing with Thomas, the main way I got sizeable negative voting was being wrong and then going to sleep without correcting it. Say 2+2 = 3 on a Thursday evening, but correct it to "1+2 = 3" a quarter hour later? Probably get one or two downvotes. Leave it that way overnight? Wake up to 10 downvotes at least. People will kick you if you don't react.
I remember picking an erroneous example in one of the endless Rust v C++ threads, and then going to bed, and yup, very clear I was wrong when I woke up.