By Guest blogger/author Naomi Hirahara
Artificial intelligence, especially ChatGPT, is on people’s minds these days. ChatGPT is developed through OpenAI, a company Elon Musk co-founded with a controlling and growing investment by Microsoft, which is beta-testing artificial intelligence in its search engine, bing. Instead of doing a simple Google search in which you type in a word, phrase, or even question, resulting in a list of search links, chat bots can provide full narratives. They are interactive, too, and can simulate a conversation with the user, albeit with mixed results. (See https://www.kpcc.org/2023-02-27/how-microsofts-experiment-in-artificial-intelligence-tech-backfired)
These developments have Google on the run and the whole high-tech community both excited and nervous about what disruptions may take place. Certain authors savvy about this world have also expressed the whole range of emotions—fear, anticipation, and indifference. Artificial intelligence is already used in narrating audiobooks for outlets like Apple Books in lieu of “real people.” (For an interesting conversation about AI audiobook narration, listen to the last story on this page: https://www.kpcc.org/show/airtalk/2023-01-23/lunar-year-massacre-in-monterey-park-leaves-ten-dead)
I’ve been listening to and reading such conversations as both a novelist and co-chair of the Imagine Little Tokyo short story contest. Regarding the latter, some journals and magazines mostly in the science-fiction realm have been recently dogged with a flood of short story submissions created on AI. (See https://www.theverge.com/2023/2/25/23613752/ai-generated-short-stories-literary-magazines-clarkesworld-science-fiction) Could someone input our guidelines on ChatGPT with locations in Little Tokyo and come up with a semblance of a good story? I can’t imagine how teachers in the future will evaluate the veracity of student essays with the spread of AI.
Joanna Penn of the Creative Penn podcast has been discussing AI for years, pointing out how writers already utilize artificial intelligence, which predicts language based on patterns, in checking our spelling and improving our prose through software programs like Grammarly and ProWritingAid. She’s also now utilizing AI exploring various creative storytelling options.
As an author who writes very specific historical and ethnic stories, I haven’t been that concerned that I can be possibly replaced by a robot. But for fun, I did go to ChatGPT and asked the bot questions about the incarceration of Japanese Americans during World War II. The answers it gave me were both encouraging and unnerving. Encouraging in that the answers were correct and factual. Unnerving because the prose was clear and well-written. Then I asked ChatGPT specifically about the topic of my recent historical mystery, CLARK AND DIVISION, which is set largely in 1944 Chicago, where many Japanese Americans released from the ten detention camps sought refuge for some months and years. For this question, ChatGPT picked up none of the historical nuance and came up with a completely wrong answer. (See screenshot.)
As I discussed recently at a Sisters in Crime Los Angeles meeting, when writing historical mysteries—or perhaps any kind of fiction—look for the gaps of knowledge. Let’s surprise our readers, take them to places and situations that they have never been. If a robot can easily replicate our tropes, characters, or plot points, maybe we should seek to create fresher tales that only living, breathing person can tell. Let’s go to those archives and open those dusty books that haven’t been digitized or engage in vulnerable, emotional conversations with people who share stories that they have never told. In this way, I see artificial intelligence as a good challenge more than a competitor. I seek to stump the robot instead of destroying it.
For a headier analysis of ChatGPT, go to science fiction writer Ted Chiang’s article in the New Yorker, in which he likens the new technology to a bad photocopy of source material: http://Ted Chiang’s Article
(Naomi Hirahara is the Edgar- and Mary Higgins Clark Award-winning author of the Mas Arai mystery series and the historical novel, Clark and Division. Evergreen, the sequel to Clark and Division, will be released this August. For more information, go to her website, http://www.naomihirahara.com.)
.
(Naomi Hirahara’s article is posted by Jackie Houchin)
Share this:
The Writers in Residence 
Insightful post, Naomi, thank you. I have a vision of the robot trying to murder itself for a grammatical error, or forgetting to recharge and collapsing into a thousand pieces then re-assembling itself to fool the reader. Perhaps a sci-fi writer has already published such a premise before being crushed to death by a jealous robot.
jill
LikeLike
If we aren’t careful, the only intelligence we will have in the world will be artificial. The prospect reminds me of the movie, The Time Machine, where all these pretty little young people do nothing and basically know nothing. (They might as well have had a cell phone in their hand like young folks do now.) In the movie they are used as food for the beasts who are running the world in the future. Not a good idea. Let machines talk to machines. I want to talk to people and write my own books for those people.
LikeLike
Naomi, thanks for your post and for being with WinR today. ChatGPT certainly concerns me, but I feel the need to be informed about it. Last week’s Washington Post has a couple of articles on the subject (I haven’t read them yet) and I look forward to checking out your links.
LikeLiked by 1 person
I was just in a webinar about potential ChatGPT uses in my day job area. Right now, I still feel like it’s a shiny toy. We’ll see.
LikeLiked by 1 person
Sharon E. Cathcart, what kind of writing do you do in your day job? Do you write fiction as well. Do you feel intimidated at all by AI and Chat GPT?
LikeLiked by 1 person
The applications of AI discussed during the meeting today had to do with medical affairs (medical information requests, data analysis, etc.). This is actually problematic because the searchable data for ChatGPT et al only goes back to 2021, which means anything it delivers is outdated the moment it arrives.
I have a second life as an author, and I’m not intimidated by AI and ChatGPT. I’m not interested in plagiarizing other creators. And that’s what it does, at the end of the day, from where I sit: enables the unethical.
As always, other people’s mileage may vary.
LikeLiked by 1 person
Thought-provoking post, Naomi, and quite interesting. I always wonder how far AI will go, especially in writing. I hope we humans can continue to get published!
LikeLiked by 1 person
Great topic, Naomi. Interestingly, I just read a NYT article about Alan Alda asking ChatGPT to write a new scene for “M*A*S*H”, which he table-read with former co-star Mike Farrell. The result: a stilted and decidedly unfunny scene. The writing we most enjoy stimulates our emotions – whether through humor, pathos or tension – in subtle and indirect ways. AI can do a great deal, but it can’t simulate what stirs us…at least now.
LikeLiked by 1 person
Reblogged this on Words and Reviews.
LikeLike
Welcome, Naomi! What an interesting view of the future! I don’t think AI has a sense of humor….. and you’re right about the nuances of historical memories that we glean from family descendants. No machine can do that. However clever it thinks it is. Bah humbug!
LikeLiked by 1 person
A very warm welcome Naomi – and thank you for such an unsettling post! I am in Rosemary’s camp with regards to the nuances and humor that no robot can replicate. But I fear that will come – perhaps not in our lifetime, but who ever thought we could fly to the moon? And yet, I’m comforted by Elizabeth Gilbert’s wonderful Ted Talk about the Creative Muse. If you haven’t listened to it, it’s really interesting. She says that the “muse” or ideas are ethereal, not tangible at all. They permeate the very air we breathe. I like to think that no computer could program the essence of a writer’s voice because that comes from our soul.
LikeLiked by 1 person
What an encouraging view, Hannah! Our muses are ethereal, not tangible. So UN-robot-like. Thanks!
LikeLiked by 1 person
Hey Writers! Here’s another post on how to use chatGPT for generating headlines and titles for you work. It’s not bad and it could help you. https://medium.com/inspired-writer/how-to-use-chatgpt-to-turn-terrible-titles-into-awesome-headlines-c677f05bc718
LikeLiked by 1 person