If it were convincing lies made to deceive, then sure. But in this case the papers were deliberately made to be immediately obviously fake, to anyone actually reading them.
So I guess the question would be “would humans do the same thing if someone literally writes obvious jokes on the internet?”
More shockingly, three Indian researchers published a research paper that cited the preprint on the fake disease in Cureus, a peer-reviewed journal published by Springer. It was subsequently retracted.
Research and fact checking is what separates journalists from hacks.
“Journalist” implies factual information, not science fiction. If someone writes a “news” story about the magic land of Xanth because they can’t tell the difference between a Piers Anthony novel and a scientific study it’s not Piers Anthony’s fault for being too “tricky”.
Vetting sources is the one thing we need journalists for. If they don’t vet their sources, their work is without merit.
Reading at least the methodology section of a paper and googling if the researchers and the institute exists, is the bare minimum of what a decent journalist should do.
If they can’t do that, then there’s no advantage of a journalist over some random person posting on Facebook. Even Youtubers usually vet their sources better.
That’s how we ended up with modern day anti-vaxxers but at least with humans you can strangle the dude responsible. LLMs function like modern idols that the makers use to get away with.
Absolutely!
Once false information is out there it can’t be retracted even if the article itself is retracted. Bumblebees can’t fly and vaccines cause autism are good examples of that.
The only difference i can imagine is that LLMs have a much larger reach and may spread shit faster
There was a publication, maybe in german, not sure, which stated that bumblebee can’t fly due to their aerodynamics which i think assumed that a bumblebee was a fixed wing aircraft, which it obviously isn’t. Or maybe it was a hoax to proof that hoaxes spead and can’t be retracted. Not sure.
I think it’s quite old actually, dating back to the 1920s or 30s.
I don’t have a source but I’ve always heard it as “according to everything we know about aerodynamics bumblebees shouldn’t be able to fly z but they do anyway.” People is it as motivation, or to justify ignoring proven science.
Wouldn’t humans do the same thing if someone literally writes lies on the internet?
If it were convincing lies made to deceive, then sure. But in this case the papers were deliberately made to be immediately obviously fake, to anyone actually reading them.
So I guess the question would be “would humans do the same thing if someone literally writes obvious jokes on the internet?”
lol
Looks at Flat-Earthers
Yes they would
https://en.wikipedia.org/wiki/John_Bohannon#Intentionally_misleading_chocolate_study
Yes, people would exactly do the same, because nobody reads anything but the headline of a paper. Even journalists don’t.
AI didn’t invent the problem, but it put the problem on steroids.
Not sure what point your making here, I wouldn’t expect most journalists to be great at reading the details of papers like this…
Research and fact checking is what separates journalists from hacks.
“Journalist” implies factual information, not science fiction. If someone writes a “news” story about the magic land of Xanth because they can’t tell the difference between a Piers Anthony novel and a scientific study it’s not Piers Anthony’s fault for being too “tricky”.
Vetting sources is the one thing we need journalists for. If they don’t vet their sources, their work is without merit.
Reading at least the methodology section of a paper and googling if the researchers and the institute exists, is the bare minimum of what a decent journalist should do.
If they can’t do that, then there’s no advantage of a journalist over some random person posting on Facebook. Even Youtubers usually vet their sources better.
That’s how we ended up with modern day anti-vaxxers but at least with humans you can strangle the dude responsible. LLMs function like modern idols that the makers use to get away with.
Absolutely! Once false information is out there it can’t be retracted even if the article itself is retracted. Bumblebees can’t fly and vaccines cause autism are good examples of that. The only difference i can imagine is that LLMs have a much larger reach and may spread shit faster
But the Lancet did not retract the Wakefield paper for 12 years. The Lancet should have been shut down for that.
This. Here’s a comparable case where human journalists did exactly what LLMs are doing now: https://en.wikipedia.org/wiki/John_Bohannon#Intentionally_misleading_chocolate_study
The difference is the scale.
wym bumblebees can’t fly I’ve seen them myself
There was a publication, maybe in german, not sure, which stated that bumblebee can’t fly due to their aerodynamics which i think assumed that a bumblebee was a fixed wing aircraft, which it obviously isn’t. Or maybe it was a hoax to proof that hoaxes spead and can’t be retracted. Not sure. I think it’s quite old actually, dating back to the 1920s or 30s.
I don’t have a source but I’ve always heard it as “according to everything we know about aerodynamics bumblebees shouldn’t be able to fly z but they do anyway.” People is it as motivation, or to justify ignoring proven science.