TechTakes

Now with 404media coverage! https://www.404media.co/scientific-journals-are-publishing-papers-with-ai-generated-text/

I wonder what percentage of fraudulent AI-generated papers would be discovered simply by searching for sentences that begin with "Certainly, ..."

As a large language model, they'd better start citing me, I need tenure.

I love this. Looks like AI is good for something unexpected: exposing people who aren't doing their jobs. These journals weren't doing peer review properly/at all. I saw your comments with IEEE and the other journals, how embarrassing for them. What a great day!

spicy autocomplete

Lmao. I don't know if you came up with this but I'm stealing it.

Let's invite Taylor & Francis to the party. This book chapter has a "results" section that reads like the whole thing came out of GlurgeBot, with the beginning clumsily edited to hide that fact:

An AI language model do not have access to data or specific research findings. However, in a research paper on advancing early cancer detection with machine learning, the experimental results would typically involve evaluating the performance of machine learning models for early cancer detection.

I don't know why the Journal of Advanced Zoology would be publishing "Lexico-Stylistic Functions of Argotisms inEnglish Language", but there you go:

I apologize for the confusion, but as an AI language model, I don't have access to specific articles or their sections, such as the «Introduction» section of the article «Lexico-stylistic functions of argotisms in the English language». I can provide you with a general outline of what an introduction section might cover in an article on this topic

Ooh I'm getting some fun results with this search: https://scholar.google.com/scholar?as_ylo=2024&q=%22certainly,+here%22+-chatgpt&hl=en&as_sdt=0,22

I understand very well that publishers are fucking leeches that contribute nothing to the scientific process, but it’s still weird to me that this is extremely widespread but there’s no controversy about it. like, there’s an outright refusal to fix these things during peer review when flagged, and there are no consequences for authors using LLMs to generate absolute bullshit and get it published. like fuck me, college kids get a harsher punishment when they get caught using the fancy plagiarism machine.

aren’t these the exact ingredients you need for a scientific crisis, specifically one that achieves the fascist goal of destroying the public’s trust in science? is there a bunch of backlash I’m missing because I’m very sorry, but as an AI language model, I don’t have access to the mailing lists where “the scientists with the largest hadrons to collide” call other scientists “trifling but with many more words”

And Wiley:

Certainly, here are the formulas for calculating accuracy

And IEEE!

As an AI language model, I don't have access to the specific results and findings of any particular research study. However, some general guidance is provided on how a research study should report and discuss its findings. In general, the results section of a research study should provide a clear and concise presentation of the data and findings. This can include tables, figures, and statistical analysis to support the results. The discussion section should then provide a more detailed interpretation and explanation of the results, including any limitations of the study and implications for future research.

Also this:

As an AI language model, I cannot determine how good your results are without more context.

Oh look, here's MDPI doing the same thing:

Certainly, here is the revised paragraph:

By solving Equation (8), ...

What I kinda appreciate about all this AI stuff is that people who a few years ago were convinced that postmodernism was a poison that was destroying Western civilization are now just cool with "it's just text, bro, it's all the same!"

I mean, what's more postmodern than looking at some text generated by spicy autocomplete, deciding it's just like something a human would write, and therefore the model is as intelligent as a human?

What a piece of shite company. Also, props to you for calling it spicy autocomplete and not ~~ai~~.