AI writing bots - what are your thoughts on their use in writing sceintific research articles?

The Verge published an article on "ChatGPT can’t be credited as an author, says world’s largest academic publisher":

“Springer Nature, the world’s largest academic publisher, has clarified its policies on the use of AI writing tools in scientific papers. The company announced this week that software like ChatGPT can’t be credited as an author in papers published in its thousands of journals. However, Springer says it has no problem with scientists using AI to help write or generate ideas for research, as long as this contribution is properly disclosed by the authors.”

 

Interesting points from the article are raised about the quality of some of the writing:

"If there is broad consensus on crediting AI as an author, though, there is less clarity on the use of AI tools to write a paper, even with proper acknowledgment. This is in part due to well-documented problems with the output of these tools. AI writing software can amplify social biases like sexism and racism and has a tendency to produce “plausible bullshit” — incorrect information presented as fact."

The recent explosion of these AI tools for writing and their apparent use already in the peer-reviewed literature was a bit of a surprise to me.  It took many years for me to improve my writing skills, and I've worked with many people to improve their skills. Good clear writing, especially scientific writing, is a mixture of knowing punctuation, grammar, vocabulary, and planning; not to mention practice. Reading well-written articles is also helpful in improving one's skills. Clarity and brevity have long been my problem - they need to be balanced to get the message across to the reader without making them read long-winded tomes.

 

Having peer-reviewed 100's of articles, I've seen the spectrum of writing - from good to bad, and have been surprised at what can slip through the publishing company's editors and made it into print.  The allure of the AI is understandable to achieve a better written outcome, but as the article notes - may frequently generate "plausible bullshit" - something we need to keep out of scientific articles.

So what is your view on using AI; whether it is for e-mails or scientific publications?