The study “Collaborating With ChatGPT: Considering the Implications of Generative Artificial Intelligence for Journalism and Media Education” by John V. Pavlik from Rutgers, The State University of New Jersey was coauthored with ChatGPT, a generative artificial intelligence platform that produces text responses from text prompts.
ChatGPT is a natural language processing platform introduced to the public in 2022 that uses OpenAI’s text interpreter called GPT-3. The letters stand for “generative pre-trained transformer”, and it is capable of reading and writing text. According to Mollman (2022), 1 million human users have signed to use it. AI, including GPT, has already been widely used in journalism, with The Guardian even publishing an article written by it. In addition to ChatGPT, there is DALL-E that creates visualizations.
The article in question was written by first providing an introduction, which I have summarized by the human part of the team and later entering queries starting with “explain how journalism works” for ChatGPT to write answers on. When reading the article, it becomes readily apparent that the answers written by ChatGPT are informative and well-written and ‘researched’.
The queries entered by the researcher include questions about journalism and AI, and the relation the two have with each other. ChatGPT is also queried about the relevance of AI to writing scripts for TV shows, to which it was sceptical as it was not designed for creative writing.
For media education in journalism, ChatGPT suggests that journalism and media educators should teach about the potentials and the limitations of AI. the upsides include being able to quickly analyze large amounts of information, automatic news generation, and improving accuracy and fairness. The downsides include the risk of introducing bias or errors, the need for careful editing, and the impact on jobs. Also, the ethical implications should be discussed.
In the concluding section, Pavlik notes that AI, ChatGPT, possesses an impressive level and range of knowledge about the topics and produces high-quality written expression that is free of errors and generally factually accurate. Of course, there are still limitations: its range and depth of knowledge and inability to think critically and creatively. The platforms may pass the Turing Test and thus may pose a threat to human professionals in regard to jobs, but it may also prove to be an useful tool especially when time and resource limitations are at play.
Suggestions for further research include critically and systematically assessing the relevance of AI platforms such as ChatGPT to journalism and media education. Educators should also consider their effect on academic integrity. There should be no delay, as the (human) author notes, the future is now.
The article “Collaborating With ChatGPT: Considering the Implications of Generative Artificial Intelligence for Journalism and Media Education” by John V. Pavlik (and ChatGPT) is in Journalism & Mass Communication Educator. (free abstract).
Picture: Untitled by Deepmind.
License Unsplash.