On october 20th, xinhua's beijing power supply went with artificial intelligenceAi(b) universalization and the use of ai tools by a growing number of scientists. A recent study published in the british journal nature human behaviour found that the scientific papers referring to ai were compared to those that did not refer to aiReference rateHigher。
This is a calligraphy smart robot filmed at the japan international robot fair, held in tokyo, japan, on 30 november 2023. Xinhua news agency's paycheck
Researchers at the northwestern university of the united states, among others, analysed about 75 million papers published between 1960 and 2019, covering 19 disciplines. It turns out that the title or summary refers to "Machine learningPapers on ai terms such as “deep neural network” are more likely to be among the top 5 per cent of the papers most frequently cited in their field and are often more frequently cited in other areas。
The study also found that over the past two decades, the institute's researchers in 19 disciplines have increased their use of ai tools, but there are significant differences, with the highest rates of use in computer science, mathematics and engineering, the lowest rates of use in history, art and politics, and geological, physics, chemistry and biology。
The study provided a quantitative analysis of how ai changed scientific research. However, in view of the survey deadlines, the study did not capture the latest developments in ai, and large-scale linguistic models such as chatgpt have changed the way some scientists study。
The results of that study also raised concerns. Lisa messery, associate professor of anthropology at yale university in the united states of america, expressed her concern that some scientists would misuse ai tools to increase the citation of their papers。
