Yeah, I'm still disappointed by that one. I was super excited by the LK99 stuff; I don't know if that was a "hoax", but it was definitely bad science that took the media by storm.
I think that the problem is that there's effectively an infinite amount of science and it changes and updates all the time, so it's impossible to be truly "caught up" with everything, and most studies are already in pretty specific niche subjects that require a lot of understanding on that niche subject. Most people doing science communication can't possibly learn it all, and most certainly aren't equipped to call out fraud of bad science in a paper, so they have to take the papers at their word.
I mean, before I dropped my PhD, I was studying formal methods in computer science. I got reasonably good with state machine models in Isabelle, so you'd think I'd be competent with "formal methods" as a concept, but not really. If I were try and read a paper on, I don't know, "Cubic Type Theory with Agda", I would have to do a lot of catching up, almost starting from scratch, and I think I'm probably better equipped than the average software engineer for that. Even if I got to a state of more-or-less understanding it, I would certainly not be equipped to call out bad science or math or fraud or anything like that.
I think that the problem is that there's effectively an infinite amount of science and it changes and updates all the time, so it's impossible to be truly "caught up" with everything, and most studies are already in pretty specific niche subjects that require a lot of understanding on that niche subject. Most people doing science communication can't possibly learn it all, and most certainly aren't equipped to call out fraud of bad science in a paper, so they have to take the papers at their word.
I mean, before I dropped my PhD, I was studying formal methods in computer science. I got reasonably good with state machine models in Isabelle, so you'd think I'd be competent with "formal methods" as a concept, but not really. If I were try and read a paper on, I don't know, "Cubic Type Theory with Agda", I would have to do a lot of catching up, almost starting from scratch, and I think I'm probably better equipped than the average software engineer for that. Even if I got to a state of more-or-less understanding it, I would certainly not be equipped to call out bad science or math or fraud or anything like that.