A Fluent, Flaky Fellow?

OpenAI's ChatGPT will write (if fees are paid) poems, essays, job applications and even scientific papers. A number of science publishers (including Springer-Nature) have now banned listing ChatGPT as a co-author on submitted articles. Other journals are even prepared to prohibit the use of the bot, as they believe it 'could pepper academic literature with flawed or even fabricated research' (https://www.theguardian.com/science/2023/jan/26/science-journals-ban-listing-of-chatgpt-as-co-author-on-papers). As a former editor of science journals, I have views the above issues. It would be wholly inappropriate to credit ChatGPT with authorship of a submitted article. The bot is technology. Would you make your spell/grammar check, use of a regular search engine or even your mobile's calculator, a co-author? Banning ChatGPT's use, however, would be a futile endeavour, especially as it's likely to appeal to authors whose first language is not English. Given ChatGPT's proven sometimes flaky outpourings, it would, however, be useful to know if it had been employed in an article's preparation. ChatGPT's output clearly conforms to the old computer adage of 'Garbage In: Garbage Out'. The 'gold standard' peer review process could well be influenced by this technology. Peers, selected by editors to review submitted articles, are not always a) fully cognisant with the latest verified data/views on very specialist topics and b) native English speakers and might be overly impressed by the submission's fluency/confidence. I guess that editors, who want to protect their journal's reputation, will just have to be a little more careful. Having said that, there are already many ways for flawed or even fabricated research to powerfully impact on human policy and thought. It's highly unlikely that peer review can protect us from that!