The arguments posited in terms of "I think it probably works THIS way ..." are arguably the best ones - they qualify someone's lack of expertise and save the reader from assigning undue confidence to claims :).
That said, Internet discussions are run on Cunningham's Law - you say what you think you know, and others will call your errors out, or challenge your assumptions. As long as people don't read any single comment as gospel, but consider the whole discussion in context of their own knowledge, applying basic critical thinking, everyone gets to train their reasoning skills and learn something. I'd expect this to be a baseline on HN.
I'm confident being the protagonist of that XKCD is a rite of passage in this industry :).
I think that there's a real problem with people using 'Cunningham's Law' as a way to learn. It might be fine for that person (because they know their limits), but then anyone reading it could incorrectly quote that person as being right (because it wasn't corrected). These ideas are then duplicated and we get into a real mess where we can't differentiate widely held views vs the experts/most agreed upon view by experts.
In general, this can't be a way to move forward as a society if people just make stuff up.
That said, Internet discussions are run on Cunningham's Law - you say what you think you know, and others will call your errors out, or challenge your assumptions. As long as people don't read any single comment as gospel, but consider the whole discussion in context of their own knowledge, applying basic critical thinking, everyone gets to train their reasoning skills and learn something. I'd expect this to be a baseline on HN.
I'm confident being the protagonist of that XKCD is a rite of passage in this industry :).