I can’t believe we’re about to throw it all away.
A couple of years ago I wrote a series of short plays about working with artificial intelligence, imagining a scenario where a company used generative AI to mass-produce TV shows, movies and podcasts. I thought it would take place several years into a dystopian future.
I was so wrong.
At the Online News Association annual conference last September in New Orleans, I learned that AI-generated podcasts are happening NOW. We are in the dystopian future I dreaded.
(Yes, yes, I know. There was plenty of other evidence — *waves arms* —before ChatGPT came along. Still.)
While I am enough of a fatalist to know that AI is part of modern life now just as much as climate change and existential dread (and yes, the three are related), it would be nice if there existed a space where we didn’t have to ask, “Did a human make this?”
That’s increasingly less true in the world of business. In social media. In school. Even in journalism. Cleveland.com’s editor said recently that the website will use AI to write first drafts of stories.
Where does it end?
I would like to think it end with theater: live performances by real people, performing words written by real people for an audience comprising real people. But even in this space, AI is creeping in. Machine-generated artwork is being used in play posters. Writers are using AI to analyze and find plot holes in their work. Theaters now are asking playwrights to stipulate that their work was not written by AI.
I am an AI skeptic, but I’m not an absolutist, believe it or not. The technology does have its uses. In my day job as a data analyst, I’ve asked Gemini to help me troubleshoot faulty code, for instance. That said, AI carries with it lots of baggage that no one, especially creators, should ignore:
- AI is not conscious and can not make decisions.
- The business model of generative AI is predicated on outright theft of intellectual property.
- The resources needed to power AI are straining the electricity grid and the water supply.
The best use case for AI can be illustrated by The Beatles, by way of Peter Jackson. The band’s final single, “Now and Then,” was brought to life using a machine-learning tool that extracted John Lennon’s vocals from a poor-quality demo. It’s a song that arguably could not have existed without AI technology, shown by the fact that it sat on a shelf for decades. The result is a lovely recording that provides a nice bookend to the Fab Four’s catalog.
When AI can do things that humans can’t do, it can be helpful. The problem is when humans ask AI to do things that we can still do.
Write.
Create.
Perform.
If we are going to relegate the quintessential tasks of human creativity to machines, what even is the point of art anymore?
I suppose I should make this about “Banshee” in some way. Here’s how AI relates to the play: In my contract with the lovely folks at Reading Theater Project, we agreed on a clause that states 1) I did not use AI in the creation of the play, and 2) RTP will not use AI in any part of the production or promotion of “Banshee.”
“Banshee” has been through about a million (give or take) rewrites since I first wrote it three years and one month ago. As recently as last week, in talks with director Jody Reppert and dramaturg Kimberly Patterson, I made some more revisions that tightened the script and improved the ending.
AI would not have been able to suggest those changes. It does not understand the human experience. So why would I even ask?
Adam-Richter.com 