“Fake abstract vs. real abstract” is less of an issue from a utilitarian standpoint.
It matters more that real and good research were done, that real and good experiments were conducted, that real and honest data were collected and published. After that, I don’t particularly care whether the researchers write the abstracts themselves or use AI to help them write the abstracts, as long as the abstracts accurately reflect the research, the experiments, and the data.
I can easily see a brilliant Japanese/Korean/Chinese researcher using AI to help him generate the English text to use to publish his work in an English language journal or symposium. It may more effectively get his ideas and results across the language barrier than the researcher doing the same without AI. And that will benefit not only the researcher but also the field of his research.
