Is ChatGPT the end of good manners?

HAVE YOU taken the Bloomberg Al-dentity Quiz? I did, and it pegged me as a “cautious optimist.” That seemed about right — but my sense of optimism was tested last week.

- By Frank Barry

No, I wasn’t reading If Anyone Builds It, Everyone Dies. Nor was I trying to determine whether a chatbot would lie, or get something wrong, or reveal its unethical underside. I was just listening to a colleague give me a tutorial on how to use one.

“Don’t be polite,” he said, as he deleted the word “please” while editing a command he was giving ChatGPT. “Saying ‘please’ just wastes its energy and resources. Be direct.”

Uh-oh, I thought. If people learn to stop using polite language when speaking to a human-like helper, won’t they also be less likely to use polite language when speaking to other people?

Courtesy and manners are habits. Once formed, they become second nature. But if using AI also becomes second nature, as the best technology eventually does, will it weaken the small courtesies that are so essential to human relationships?

After ChatGPT delivered a first draft, my AI tutor ended a long followup prompt by writing: “DO BETTER. YOU CAN DO A LOT BETTER.”

Wow, I thought — that’s aggressive. Chatbots are often compared to highly capable interns who need specific guidance, but this felt more like we were dog trainers scolding a puppy or football coaches haranguing a linebacker who missed a tackle.