For The Boston Globe, I wrote an article on why we should accept authors using AI not only for tasks like researching and spell check, but also for composing text. As always, you can read it by subscribing to the Boston Globe newspaper or being a paid subscriber here.
Should writers use AI? It’s a topic that has generated frenzied discussion. Recently, Washington Post columnist Megan McArdle kicked up a storm when she wrote on X that she uses AI for such tasks as transcribing interviews, analyzing her arguments, and fact-checking. Critics accused her of outsourcing her thinking and told her to find a new job. A Rutgers philosophy professor even said that “in a healthier media culture, an admission like this would at the very least get her fired.” One wonders what he means by at the very least and what kind of punishment he has in mind as a maximum.
Yet the arguments one can make against her use of AI also apply to other forms of technology that are widely considered acceptable. In fact, I would go one step further than McArdle and say that there is nothing inherently wrong with writers using AI to compose text. The only way it would be unethical is if writers fail to disclose it, because people have an obligation to disclose information that publishers or readers might find relevant. Since I can write fast and well enough not to benefit much from AI writing, and because it would be against the rules of Globe Ideas anyway, I did not use AI to write this piece. But I would encourage publications to avoid blanket rules against such a practice for those who need it. (I did use AI to check spelling and grammar before submitting my draft.)
