The Practical Testing,Analysis,and Recommendations of ChatGPT in the Publication of Sci-Tech Journals—A Case Study of Physics Related Sci-Tech Papers
In order to investigate the potential impact of artificial intelligence(AI)on the publication of sci-tech journals,this article takes the participation of the Chat Generative Pre-trained Transformer(ChatGPT)in the entire process of writing and publishing physics related sci-tech papers as an example,and studies the roles and limitations of ChatGPT and other tools in literature searching,literature reading,formula derivation,data processing,paper writing,column topic planning and solicitation of contributions,plagiarism checking,editorial quality inspection and so on.It finds that AI can replace humans in performing low-creativity and repetitive tasks in the publication process of sci-tech journals,but its role is limited in complex and innovative work.AI has brought three opportunities to the publication of sci-tech journals:liberating productivity to accelerate knowledge innovation,showcasing technological advantages for improving publication quality,and enhancing the efficiency of publishing content dissemination to accelerate the technological progress.It also poses challenges such as leakage risks,limited functionality,and user reliance on tools.In view of this,it is necessary to help AI to better play its role in the publication of sci-tech journals by improving the rule system and optimizing the regulatory mechanisms to strictly control the occurrence of leaks,accelerating the research and development of AI based on Chinese natural language to seize the technological discourse power,strengthening knowledge updates and promoting the organic integration of human-machine interaction to avoide the reliance risks,and strengthening research to clarify the tool attribute of AI and reasonably define user responsibilities.
artificial intelligence(AI)Chat Generative Pre-trained Transformer(ChatGPT)publication of sci-tech journalsknowledge innovationtechnological discourse power
NETL
NSTL
万方数据
