Le Bonheur Children's Hospital fellow publishes journal article on chatbots written by ChatGPT

MEMPHIS, Tenn., April 24, 2023 /PRNewswire/ -- The use of chatbots and natural language processing technology, such as ChatGPT, has the potential to change the future of medical writing —with some caveats— says Som Biswas, MD, a Radiology fellow at Le Bonheur Children's Hospital and the University of Tennessee Health Science Center, in an article published in Radiology. The published article was written entirely by ChatGPT with editing from Biswas.

ChatGPT is an artificial intelligence model created by the research company Open AI. Launched in November 2022, ChatGPT and others like it generate text by processing large datasets of text and producing similar text. It uses natural language processing (NLP) technology, a field of computer science that develops systems to understand and generate human language. These technologies have many potential uses in the field of medical writing, says Biswas. The top benefits include automation and efficiency, such as extracting information from electronic medical records or assisting with literature searches, and improving the speed and accuracy of document creation, such as article generation and meeting summaries.

"The use of chatbots and NLP technology has the potential to change the way medical writers operate," said Biswas. "Documents could be more accurate, more consistent and require less effort from a human medical writer."

The article stipulates that these technologies need to evolve to be effective in the medical field and will not replace human medical writers in the near future. First, language models do not have the medical expertise and specialized knowledge needed to write in depth on particular subjects. Second, ethical issues regarding plagiarism, bias and errors are a concern when using the current version of ChatGPT.

Biswas noted multiple cautions to consider when using chatbots in the medical writing process:

• Ethics: Potential for plagiarism and concerns about accountability
• Legal issues: Copyright, compliance and other medico-legal considerations
• Innovation: Current technology leads to repetitive text generation and lack of creativity
• Accuracy: Potential for AI-generated text to contain errors
• Bias: May perpetuate bias included in the data on which they are trained
• Transparency: Clarity when AI is used in the writing process

"Overall AI-powered language models are powerful tools that could assist human writers in some tasks," said Biswas. "However, as we move forward with adopting this technology, we must consider the limitations and potential risks of using AI in the writing process."

SOURCE Le Bonheur Children's Hospital