zlacker

[parent] [thread] 8 comments
1. bgoate+(OP)[view] [source] 2024-06-07 15:07:38
Having been involved in peer reviewed publishing before, I wonder if this was an afterthought prompted by a peer reviewer's comments on the paper. Perhaps they quickly added this point just to get it to pass review. Sloppy, if so, but I've seen similar (though not as blatant) things happen.
replies(2): >>crdros+85 >>karate+ui
2. crdros+85[view] [source] 2024-06-07 15:44:15
>>bgoate+(OP)
I mean the whole paper is sloppy here. They kind of are writing to a biomedical journal and then say "well, it's NOT biomedical..." as kind of a little brush-off? And then they go through a couple explanations they have heard and dismiss them without really going through the evidence. (I especially liked the sloppiness of self-contradiction, in one section they're like "well there are no Hebrew letters that work for this" which is wrong, you could make decent arguments for both shin and tsadeh -- but then almost immediately after they're like "well this could be an M or W, W can symbolically be the Hebrew letter Vav...")

And after cursorily dismissing them they just say "therefore, it's an aesthetic meme. This is just what perfect hands look like, sorry."

A hypothesis that was not considered, for example: 'We asked people at school to imagine that they were going to be sitting still for the next three hours on a stool, and to sit on it in a way that was perfectly relaxed. We then prompted them "remember, in old times you'd have had to sit here for three hours, really relax." Finally, we then picked up their left wrist, turned it, and placed it on their chests saying "great, now can you just hold this hand here," and took a photograph. In 30% of these photographs we also see, even without syndactyly, that the two fingers get forced together just by the process of having your wrist twisted by an artist and then the fingers having to conform to the contours of the chest.'

I don't know what that percentage is, but I'd be surprised if it were 0%, right?

replies(1): >>burnis+jo
3. karate+ui[view] [source] 2024-06-07 17:01:51
>>bgoate+(OP)
There were enough basic grammatical errors in that article—not to mention a general lack of clarity and specificity—that I initially wondered whether it was a preprint, or maybe somebody's blog. But no.
replies(1): >>standa+cd1
◧◩
4. burnis+jo[view] [source] [discussion] 2024-06-07 17:39:22
>>crdros+85
The genius of your proposition is that you wouldn't have to be surprised, you'd just know
◧◩
5. standa+cd1[view] [source] [discussion] 2024-06-07 23:51:55
>>karate+ui
"According to this hypothesis, the gesture was a secret sign used to recognize crypto-jews each other"

"According to this hypothesis, the gesture was a secret sign used to recognize masonic followers each other"

I have never heard this verbiage before... Did an AI write this? Or, can someone explain to me how "used to recognize followers each other" is grammatically sound?

replies(2): >>mkl+4x1 >>karate+uj2
◧◩◪
6. mkl+4x1[view] [source] [discussion] 2024-06-08 04:37:31
>>standa+cd1
It's not grammatical. Reading the article parts of it seemed poorly machine-translated to me (and the whole thing seemed mostly a sequence of straw-men).
replies(1): >>b112+AI1
◧◩◪◨
7. b112+AI1[view] [source] [discussion] 2024-06-08 07:53:02
>>mkl+4x1
And yet...

I recall when young, people were commenting on how most media came from centralized locations. That with newspapers, and then radio, and now TV, pronunciation was moving towards being less regionalized, diverse, yet also that the choice of words to use, the synonyms to use, was changing.

I also recall the same being said for a variety of things, such as spell checkers, and grammar checkers used in wordprocessors. Some grammar was "OK", but other forms were being pushed by (most especially) earlier wordprocessors, with grammatically valid text being marked with that wavy underline.

Now we have AI.

My point?

Kids are going to be raised in a world with AI. If it spends a decade or more spewing blather such as this, an entire slew of people will grow up, from 10 to 20 years old, 15 to 25 years old, learning to cobble together sentences in this sort of way.

Not only will they read it, but "helpful" assistants will change their normal prose, into this gibberish.

So I'm sorry mkl, it sort of will be grammatical. And no, I'm not happy about it.

replies(1): >>Hnrobe+7Z1
◧◩◪◨⬒
8. Hnrobe+7Z1[view] [source] [discussion] 2024-06-08 12:18:35
>>b112+AI1
Indeed. And now imagine an opinionated government controls the AI which helps formulate not only articles, blogs, etc, but even the predictive text software that facilitates writing a language very poorly suited to writing on a phone. And then imagine how much harder it will be to overthrow that government when even your own speech is subtly nudged to support the will of Big Brother. And there you have the fate of the Chinese people.
◧◩◪
9. karate+uj2[view] [source] [discussion] 2024-06-08 15:47:20
>>standa+cd1
> Did an AI write this?

I think you know an AI didn't generate the sentence because it's ungrammatical.

The publication in question (Acta Biomed) is oriented around "mainly national and international scientific activities from Mediterranean and Middle Eastern countries", so there's a reasonable chance none of the the authors speak English as a first language. This in no way excuses bad editing—the journal publishes in English, after all.

[go to top]