[exmadscientist]

I don't know what to tell you, man. It's classic ChatGPT output, with its weird italics, sometimes-bolded bullet point headers, oddly placed and oddly frequent em dashes, and generally really distinct voice. I didn't recognize it until I started to use ChatGPT myself, and now I see it everywhere.

I also distrust it immediately, because I know how often ChatGPT bullshits me, so I can't help but assume it's bullshitting here too.

[MichaelBurjack]

You keep attacking the layout and formatting of the article, and not the substance.

Maybe this article works better for you, and if not, I'm sure you're just as capable at using Google as I am. There are many other high-quality studies that cover this topic in exhaustive detail.

https://www.cidrap.umn.edu/covid-19/commentary-wear-respirat...

[anonymouskimmer]

As a novid, thanks for taking the time to educate here.

[wizzwizz4]

This is an FAQ where each entry has a TL;DR. For question 9 in particular, the list consists of items and explanation, where the author chose to use <ul> / <strong> instead of <dl> / <dt> / <dd>. This is one of the situations where the "sometimes-bolded bullet point headers" formatting is appropriate. (The most semantically-correct formatting would be paragraph headings, as seen in LaTeX; but HTML doesn't have these.)

The <em> tag is used to indicate stress emphasis. This is the intended purpose for which the tag was added to HTML, not "weird italics". (I type by transcribing my speech, so I tend to overuse this: one of my editing passes is removing unnecessary <em>s.) This article only contains 9 <em>s in 10 questions: of these, I'd remove the emphasis from two of three "well-fitted masks", and reduce the other to just "well-fitted".

Unspaced em-dashes are often used to offset parentheticals – though I prefer spaced en-dashes myself – and these are both long-standing conventions (see https://en.wiktionary.org/wiki/%E2%80%94). Parenthetical dashes are common in formal writing, and this is formal writing.

As someone who frequently wrote in more-or-less this style (where appropriate) before GPT-1 was even made, who's also fairly decent at spotting ChatGPT output, I don't think this is ChatGPT at all. Apart from superficial formatting considerations, it's not the distinctive ChatGPT voice; and the most distinctive part of ChatGPT output is its inappropriate use of voice and formatting, whereas all of these stylistic choices are easily-justified. Perhaps most importantly, it actually says something.