Point taken: ‘This technology is very useful, it's also kind of terrifying’

Thousands of AI developers recently wrote an open letter calling for a pause on AI development — and regulation for the very technology they’re creating.

On Point has closely reported the revolution in artificial intelligence with ChatGPT, our award-winning series Smarter health: Artificial intelligence and the future of American health care and most recently in an episode that asked: Can AI be regulated?

For our episode on AI regulation, we asked listeners how AI appears in their workplaces and lives, and what guardrails or regulations they might like to see in the future.

Here’s what On Point listeners shared:

Jennifer Lapham, Orrington, ME

“Is AI being used as a tool to preserve languages that are at risk of no longer being used?” is a question listener Jennifer left for us in this voicemail:

"There's certainly a lot of focus on possible, destructive, dangerous capacity. But as the conversation focuses on how many words AI can learn and store and produce. I wondered if there is any work going on to use AI to preserve languages that … I don't want to use the term dying languages.

"But languages that are spoken only by small numbers of people at this point and are at risk of no longer being live languages. So I know there are institutes that are doing this work but can AI be used to help preserve all of the vast variety and richness of human language?"

Bruce Kaplan, Richmond, CA

Bruce Kaplan is a musician and graphic designer who has experimented with a few different AI platforms. Recently, he asked AI platform Midjourney to make a poster for a music festival.

"I used Midjourney to make a series of images of imaginary bands that were made up of animals. And the results were really quite astoundingly good. It probably saved me $1,000 in illustration fees, and it probably cut ten days off of the process by not having to hire an illustrator.

"Though this technology is very useful, it's also kind of terrifying, and I'm nervous about what nefarious people will do with it. And I'm less than confident that the government will be ahead of the curve in regulating it."

Amanda Elend, Clearwater, FL

Amanda from Clearwater Florida tried using ChatGPT for a writing class she’s teaching on voice.

She asked it to create an example of someone writing in the style of author Toni Morrison. She starts off her message by reading what ChatGPT came up with:

"‘The morning that [she] lost her voice was like any other. The sun was just beginning to rise above the trees, casting a soft golden glow over the fields. She woke up to the sound of birds singing outside her window.

'Their melodic chirps filling the room. But as she opened her mouth to let out a yawn, she felt a sudden pain in her throat. It was as if something had lodged itself there, preventing her from speaking.’

"Okay, so we all know that's not Toni Morrison, but it definitely has some of the characteristics of Toni Morrison, so it feels like the perfect illustration of voice. If somebody says it's Toni Morrison, it has her voice. So that's a little concerning."