It’s hard to think of an industry that isn’t worried about how to deal with the explosion of artificial intelligence programs but I’m not sure any are more perplexed than writers, especially writers of fiction.

“We don’t really know yet what is right action. It makes the whole process of writing more difficult than it was before,” is how James Patrick Kelly put it during a big writers conference Saturday in Concord. “We’re wondering, can I do this? Are the A.I. cops looking over my shoulder on this?”

If an experienced pro like Kelly is uncertain, then every writer should be.

Kelley, who lives in New Hampshire, has had the career I wanted when I was 17: He has been publishing science novels and short stories for 50 years; has won both top awards in his field, the Hugo and Nebula; writes a regular column in the field’s most prestigious magazine (which, incidentally, rejected my youthful effort at writing sci-fi); has people paying for his advice on writing and publishing, including teaching a Master of Fine Arts course for many years; and is still going strong.

That reputation is why he was invited by the New Hampshire Writers Project to be one of three published writers making up a panel discussion titled “A.I. in the Literary World.” The panel, which I moderated, was part of the day-long 603 Writers Conference at the Grappone Center, which drew some 200 writers, editors and wannabe writers. Judging from the packed hall for the panel there is plenty of angst in the literary world.

Kelly argues that writers shouldn’t approach A.I., meaning mostly large language models like ChatGPT, with an all-or-nothing attitude. It is too useful a tool in certain areas. The complication, he said, is that publishing wants authors to avoid the technology altogether when it is changing at light speed and infiltrating everything.

“The fiction industry has made its mind up about what is too far: and ‘too far’ means there should be no words in your manuscript that were written not by you, that were suggested by A.I.,” he said.

That sounds reasonable, of course. I want to read a human’s thoughts when I read a story, not the product of an probabilistic algorithm.

But Kelly gave the example of doing research for the story, in this case looking up a description of St. Mark’s Square in Venice. These days it’s probably an A.I. “summary” of online material.

“So how many words in that description can you use? Can you use any of them?” he said. “But I’ve seen that passage, so I have to forget that I’ve seen it and try not to use any of those words. I don’t know what part of that research is actually kosher to use and I need to worry about that, which I don’t need to worry about before.”

Of course at the start of Kelly’s career he could have looked up a description of St. Mark’s Square in an encyclopedia and still have the same “how much of this can I use?” quandary. But back then editors or online influencers couldn’t have run his work through encyclopedia detector software and brought down outrage if they thought they had found something. Today that’s a real possibility.

Panelist Dena Stahlheber put forward one approach for a working writer, especially when doing nonfiction such as her book “The Heart of Prayer” on Bible-based approach to finding God. She said she documents what types of A.I. she used and for what purpose, then includes that material in submissions. “You let people know,” she said.

However the other panelist, Indie speculative-fiction writer Jeff Deck, was having none of it. He said he avoids A.I. like the plague partly because he considers it something of a plague due to its grotesque environmental cost and the nasty habits of the companies that are shoving it down our throats.

On a more down-to-earth level everybody, including questions from the audience, showed alarm about what “A.I. slop” is doing to publishing. People use it to spew out vast reams of books and articles that aren’t good but aren’t god-awful and are getting less awful by the month, cluttering up the industry and making it harder for authors to get published.

Kelly gave the room a downer of an answer when one audience member asked if readers really care whether something has been written by an A.I. or a human as long as it meets their needs. His response: No, they don’t. That isn’t what people wanted to hear.

Finally, at the deepest level, there was dismay mixed with bafflement about the whole idea of using A.I. to craft your prose. The whole point of writing is to express what it means to be human, not just for the reader but also for the author. Getting software to do it for you is missing the whole point, like going to the gym to build muscles and then using a forklift to pick up the barbells.

As I said at the start, plenty of professions are having similar discussions. If you want angst and confusion about A.I., talk to a theoretical mathematician; they’ve seen 2,500 years of intellectual practice heaved out the window overnight.

But the possibility that A.I. could undermine the millennia-long practice of people telling each other stories is even more distressing. Storytelling is an inherent part of being human; if we lose it, we lose a part of – dare I say it? – our soul. That seems like something to push back on.

For those who are wondering, I and other Monitor reporters use A.I. software to suggest headlines and summaries that we can adopt or edit as we wish, but not to write or edit the stories themselves. Aside from spellcheck, of course; how did we live without spellcheck?

this sort of guilty – can I do this, are the AI cops loking over my shoulder on this?

we don’t really know yet, what is right action

it makes the whole process of writing more difficult than it was before

writing about a scene … if you pull up a thing, this is what

how many words in that description can you use? can you sue any of them?

but I’ve seen that passage – I have to forget that I’ve seen it and try not to use any of those words

don’t know what part of that research is actually kosher to use

I need to worry about that, which I don’t need to worry about before

this panel features author and executive leader Dena Stahlheber, award-winning science fiction author James Patrick Kelly, who brings a unique perspective on emerging technology, and indie author, writing coach, and editor Jeff Deck.

David Brooks can be reached at dbrooks@cmonitor.com. Sign up for his Granite Geek weekly email newsletter at granitegeek.org.