icon caret-left icon caret-right instagram pinterest linkedin facebook twitter goodreads question-circle facebook circle twitter circle linkedin circle instagram circle goodreads circle pinterest circle

One Writer's World

Listening to an AI voice

This morning life became a bit surreal for me. You may have heard about ChatGPT; this is AI writing books and essays well enough to fool people and let students get A grades without any effort. Education on writing and thinking will be replaced, along with writers who create novels and research articles. At least that's the fear.
You and I may fear our writing skills and imaginations are no longer enough to compete in the literary marketplace but this morning I learned that it gets worse: we may be displaced in another realm—interviews on a podcast. On the CyberWire website, Brandon Karpf interviews AI, after it has been given a voice (with a British accent), and then discusses the interview with Rick Howard. The IT voice sounds reasonable and polite all the way through, even able to explore the problems with AI, which the two humans discuss at greater length at the end of the interview with AI.
I'll admit that I was slow to adopt computer technology because I found the earlier computers frustrating to use. Typing on a typewriter with paper was easier. Also in the beginning I read short and long pieces that had been sent out into the world before they were ready, in my view. Getting something finished so easily on a computer tended to lead the author to believe that the essay or book was actually finished. The pages of an essay never looked finished on a typewriter until the work had been fully rewritten and polished, but that stage could easily be ignored on a computer. Writers have dealt with this temptation and now know to question the assumption that something is finished because the page looks so perfect. But AI now offers greater temptations.
Some of this is funny. The AI voice could offer "advice" to the writers on how to correct the weaknesses in an essay. How easy would it be to slip in all sorts of material to undermine—what?—facts, confidence, conclusions? I don't know, but here is the interview for those of you ready to hear a software (?) talk.
You can listen to the podcast here.

Be the first to comment