AI: Confessions of a reluctant convert and the fear that remains
Two years ago I was at the Traverse travel bloggers conference in Tbilisi, Georgia, worrying out loud about my upcoming memoir that had no title. My editor had come up with, One Way Ticket to Rome.
It was the best we’d come up with – until I did a Google search on whether other books had been called One Way Ticket to …
I stopped counting at 50.
Traverse consists of a weekend of highly defined sessions designed to help every aspect of the booming travel writing business. Max Hartshorne, my old editor at GoNomad.com, one of the top five travel websites in the world, pointed to the session on AI. Artificial Intelligence.
In my book, AI was a four-letter word. It was a computer designed to destroy jobs, lower salaries and force feed society with a steady stream of literary mediocrity.
“Try it,” Max said. “You might get some ideas.”
The speaker spent an hour listing all the ways AI can help society, not hurt. One suggestion hit home: book titles.
I rushed out of the meeting and sat down. I immediately subscribed to ChatGPT, one of the many AI forms the speaker suggested. I typed “Book title for memoir about a travel journalist who lived a decade in Rome.”
Before I could check my email, ChatGPT spat out about 40 titles. I stopped at the first one on top.
That wasn’t mediocre. That was perfect. I added the subhead, An American Journalist’s Decade in Rome and published it (above) four months later.
Call me a reluctant AI convert. It has become my guilty pleasure. AI is like Amazon. You know it’s destroying businesses and jobs but it’s so convenient, you can’t help but use it.
May the literary gods not smote me to dust.
How I use AI
I find myself using AI every day, eating guilt with every question I type in ChatGPT’s little box. Is it creeping laziness or am I finally seeing AI’s value, kind of like the matronly woman who treats you so kindly you fall for her?
Just in recent weeks, this is how AI has helped my life:
- Before visiting Krakow three weeks ago, I asked ChatGPT, “Give me a story angle about Auschwitz that is different from the normal, ‘Nazis were bad.’” It gave me about a half dozen ideas. One was: Interview the tour guide. What’s it like psychologically spending every day showing the world the most terrifying place in history? My guide was great and I loved the blog it produced.
- I’m going to Kurdistan in September. It’s a self-autonomous region in Iraq and is promoting tourism in this period of peace. I asked ChatGPT for a three-day itinerary. It gave the capital of Erbil, the lakes and mountains of Suleymaniye and Ramanduz, famous for its craters and waterfalls. ChatGPT included photos, handy tips and time schedules.
- On July 1, Marina and I leave for our annual trip to the Greek isle of Skopelos. I’ve been to this island of 5,000 people six times. I write a travel blog on every trip. I had exhausted all ideas for this year. I was stumped. How much can I write about feta? I asked ChatGPT. It said explore the myths and fairy tales that have circulated on the island for generations. I asked our learned manager at the Panormos Beach Hotel the same question. He answered: the island’s myths and fairy tales.
- I am doing a new ad campaign for The Cappuccino Chronicles, complete with a video ad by this superb video genius in the Philippines, Kresja Maryn. She is using AI for the video montage. My publicist, April Cox, is using AI to create a new cover.
The problems with AI
Then again, AI has an evil side, the one intent to burn society’s literary soul. I have never used Chat GPT to write an article. Never will. But I’ve read stories of college students who get entire term papers written in about the time I took to roll a piece of paper into my old 1929 Royal typewriter in college.
I’ve seen it with my website.
Dog-Eared Passport occasionally run guest blogs. I get offers every day but only run about one a month. It’s vino money. That’s it.
Guest blogs are travel blogs cleverly designed to include ad links to products the pitching agency’s clients are selling. If the story angle is good, I’ll run it.
Sometimes it’s too good.
One time an agent in India sent me an article about the pitfalls of traveling alone in Paris. It was fine. But something felt off. It didn’t have a single missed punctuation or misspelled word, not even a typo.
It was totally neutral. No opinions. Not great writing but not bad, either. It had bullet points, a summary at the end.
It had the lilt of a woman. I asked for a byline. He wouldn’t give it to me. Finally, after the fourth demand he sent me a photo and byline of some guy in his office. This did not add up.
I went to – you guessed it – ChatGPT. I asked “What are the common characteristics of an article written by AI?” It said neutral prose, bullet points, summary at the end.
The Indian agent was flummoxed that I asked if AI wrote it. Yes, he said. So? He said about 90 percent of all guest blogs are AI written. I turned him down and he flipped out, casting Hindu curses on me and writing that AI – and he – predicts I’ll die a painful death.
Sorry. AI can’t predict the future.
At least, not yet.
The numbers
The numbers are staggering. Industry analysts claim 500-600 million people use AI every day. Between 1.7-1.8 billion have used it at least once. That’s one of every five to seven people on Earth. Nineteen percent of adults use it in the U.S.
I am one of 1 billion people who use ChatGPT every month.
Yet the negatives are numerous:
- The World Economic Forum reported that by 2030, AI will displace 92 million jobs.
- The amount of private data AI can analyze raises major questions about our right to privacy.
- I’ve caught errors on ChatGPT. That means misinformation can spread worldwide.
I have already seen the damage it can do to my beloved newspaper industry. Gannett Newspapers started using AI to write high school sports reports in the Columbus Dispatch, Louisville Courier-Journal and Milwaukee Journal-Sentinel.
I cut my teeth in college covering high school sports for my hometown Eugene (Ore.) Register-Guard. It’s where I drew praise from the veteran writers I idolized when I called Danny Ainge-led North Eugene High’s basketball powerhouse, “The greatest slaughterhouse this side of the Chicago stockyards.”
AI couldn’t come up with that.
Then I read Gannett merely fed a stat sheet into a computer and spat out a game story before their reporter could write his byline.
Gannett, however, ended the practice when readers protested.
Publications have sued AI for copyright infringement after their information popped up in their answers.
But then, AI converts are growing like a religious cult. The World Economic Forum also reported that by 2030, AI will recreate 170 million new jobs, nearly 80 percent more than were lost. (AI, in its own defense, said the jobs aren’t really lost as those people replaced by AI can retrain and return to work – in AI.)
The AI debate
The war of words is gaining steam. On my Facebook feed, I saw Sam Altman, the CEO of OpenAI, saying, “We are deploying responsibly developed AI to liberate humanity from mental labor and usher in an era of unparalleled creativity and abundance.”
Shot back Canadian author Naomi Klein: “You ingested the entire written output of human civilization without consent, without compensation, and without credit, to build a system whose primary commercial application is eliminating the jobs of the people whose work you consumed. You are not liberating human creativity – you are strip.mining it and selling it back at a markup while calling the theft ‘training data.’”
Gary Marcus, an NYU professor, cognitive scientist and major AI critic, told the MIT Technology Review, “I’m depressed about it. When I went into this field, it was not so that we could have a massive turnover of wealth from artists to big corporations. The people who are running AI don’t really care that much about what you might call responsible AI, and that the consequences for society may be severe.”
I fear some day that the world’s stadium press boxes will stand empty, sans some gnome feeding facts, figures and quotes into a computer. I fear brainstorming sessions in bars will be replaced by questions typed into an AI box.
Yet I feel I’m getting sucked in. I’m twirling in a whirlpool of technology that will suck me to depths where I can’t see reality with my own eyes. A computer will do it for me. Why do I worry? See the headline above?
I got it from AI.