Abstracts written by ChatGPT fool scientists

By Holly Else. Nature, 2023

542 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Health Informatics

View Paper: https://www.semanticscholar.org/paper/50aea7bae4c478e850f218e58da0e24f501ab8fc

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Health Informatics

Citations in the corpus

Can artificial intelligence help for scientific writing? (2023, background): “Moreover, it may be difficult to recognize whether a paper is written by a chatbot or a human being [5], because chatbots use advanced techniques, such as natural language processing (NLP) and machine learning, to generate text that is similar to human writing.”
From human writing to artificial intelligence generated text: examining the prospects and potential threats of ChatGPT in academic writing (2023, unknown): “Indeed, and As an illustration of the latter [18], we asked ChatGPT to generate the abstract of the present manuscript (Box 1).”
ChatGPT and the rise of large language models: the new AI-driven infodemic threat in public health (2023, background): “In fields where fake information can endanger people’s safety, such as medicine, journals may have to take a more rigorous approach to verify the information as accurate (53).”
ChatGPT in Healthcare: A Taxonomy and Systematic Review (2023, background): “However, banning ChatGPT from scientific writing is not easily enforceable, since ChatGPT is trained to produce human-like texts that even scientists and specifically-trained AI detector sometimes fail to detect [29, 7].”
Artificial Hallucinations in ChatGPT: Implications in Scientific Writing (2023, background): “Interestingly, the human reviewers stated that it was difficult to identify whether the abstract was written by a human author or a chatbot [1,2].”
Large language models in medicine (2023, background): “On the other hand, explainable AI initiatives that address the black box issue and facilitate deeper understanding of bias and other ethical issues could have benefits beyond LLM applications, by providing new investigational approaches and insights into linguistic processing in the human brain 87 .”
ChatGPT in medicine: an overview of its applications, advantages, limitations, future prospects, and ethical considerations (2023, background): “It was found that ChatGPT-generated abstracts smoothed through the plagiarism detection software: the median originality score was 100%, showing that no plagiarism was detected while the AI-output checker identiﬁed only 66% of the generated abstracts (Else, 2023).”