64 thoughts on “Are Writers Obsolete Yet?”

1. Thanks for the deep dive into AI software, Debbie. I tried one of those programs as a prompt. I didn’t care for it. It gave me a bit too much, to the extent that I felt like I was being led down a cattle chute. Nope.

   Have a great week!

   • Joe, being led down a cattle chute is an apt description. I wonder how much we’re giving up in the pursuit of convenience and instant gratification.

     There’s a difference between intelligence and wisdom. Your wise brain will produce better writing than any software can.

     • Wisdom is intelligence polished by experience… something AI will (at least now), never have…

       • True words, George.

2. I have a program called Sudowrite. I use it when I’m stumped on how to describe a place or how to show deeper emotion. I hope it’s maki g me a better writer, not a lazy one.

   • Jane, Sudowrite came up often in my research as a useful tool. As long as your goal is to become a better writer, I doubt you’ll ever be a lazy one!

3. Thanks, Debbie. Excellent summary of the many issues emerging from increased use of AI. As the CNET example showed, readers are at risk of receiving inaccurate information, a real-time train wreck with toxic side effects.

   The manipulation of readers has continued since the invention of the printing press. What’s different today is the speed at which the manipulation takes place and the number of digital channels the manipulators can distribute and exploit the public.

   AI can and is exacerbating this manipulation.

   For an analysis of the manipulation, Ryan Holiday’s TRUST ME I’M LYING: CONFESSIONS OF A MEDIA MANIPULATOR (5th Anniversary Edition) lists dozens. As one commenter wrote, TRUST ME is “A playbook for the dark arts of exploiting the media.” And when manipulators exploit the media, they influence the public, often in the worst possible ways.

   Add AI to the mix and you have the potential for an all out challenge to the way we interpret what is real, making it impossible to discern facts from manipulation.

   Even more frightening is how the mainstream media picks up and publishes distorted posts, turning a fringe thought into what becomes an accepted belief.

   If someone tries to confuse the press with the facts, they almost never publish a retraction, and if they do, it’s never seen because of the influx of AI-generated articles.

   AI generated traffic appeals to schemers, and they’ll use it with no restraints to produce more revenue.

   AI written books and articles are not just alarming, but 1984-scary.

   • Grant, thanks for joining the discussion and adding your articulate, thoughtful analysis. The gee-whiz excitement of new technology often ignores the long-term effects of that technology.

   • I’m glad you see ahead to where this is going. It’s a disaster on the horizon, with nothing to stop it, another instance of a slow-train-wreck in our world.

4. I’m an un-repentent curmudgeon on such things. I’m all for advancements, but not into the “shiny new object” syndrome that so many have about such things as this—that ‘go overboard’ mentality to latch onto stuff like this without thinking about the true ramifications. And do people think for some reason that if they use AI it won’t be error prone? How can this be since humans created it? And that gets to the crux of every matter–the condition of the human heart and mind–it affects everything we do. Which is why we should always be thinking for ourselves.

   You bring up an interesting thing that I hadn’t even begun to consider–an uptick in submissions due to AI generated content. I didn’t realize the impact was already being felt so strongly. I personally can’t imagine using AI for writing–I mean where’s the fun in that? Not even for a school paper. If I’m going to school, I’m going to learn and explore. Having someone or something else do my exploring for me is pointless. Why bother to waste the high cost of tuition?

   What will society learn from this new wrinkle? Guess we’ll just have to wait and see…

   • Brenda, closing submissions was news to me, also, which is why I wrote this post. “Thinkiing for ourselves” is increasingly rare.

     Glad to include you into the curmudgeon’s club.

5. Seems to me author branding will become more important for us human scribes. Can AI duplicate the warmth of communicating with readers via newsletters and the like? Or on blogs? I think not. About all it will be able to say at the end of a “novel” is, “I’ll be back.”

   • Jim, you nailed it–communiicating with readers is what writing is all about. If I want to interact with a machine, I’ll use an ATM.

   • Also, if I may, on the topic of branding…

     Author written works (I doubt “Author Intelligence” will be referred to as “AI+”), could be marketed as “Organically Written” in much the same way produce is sold as “Organically Grown”…

     …which raises the question: Do word processors process words in the same way packaged foods are called processed food? Under AI, I think the answer to that question can be deduced…

     • George, as I wrote in my own blog post awhile back, beginning with my next publication, the following statement (or something similar) will go into the front matter:

       This fiction is a Creation, the result of a partnership between a human writer and the character(s) he accessed with his creative subconscious. This is in no part the block-by-block, artificial Construction of any sort of so-called artificial intelligence. What you read here is what actually happened there.

       • Harvey, I like your disclaimer. Thanks!

         • You’re welcome, Deb. Everyone should feel free to use it or alter it for their own use.

       • Great, Harvey! Good thought.

         • Thanks, Steve.

6. Wow. A lot to absorb first thing in the morning. Thanks for your research.
   I don’t think AI is close to capturing the emotion of fiction, the connection to characters. If the publishing houses can recognize AI generated works, then there’s a long way to go. Garry’s demo of an AI generated blog post was recognizably “off” to his readers.

   • Terry, at this point, AI-generated content is still recognizable as such in many cases, like Garry’s post. But the sophistication is improving, making it more difficult to identify.

     Here’s an interesting article that talks about the difference between artificial intelligence and human intelligence: https://www.upgrad.com/blog/ai-vs-human-intelligence/#:~:text=Humans%20use%20the%20brain's%20computing,and%20gets%20accustomed%20to%20them.

7. Great post, Debbie. Eye opening. I think emotion is related to the soul, and AI will never have a soul.

   I like your ChatGPT spell check with the result “Cat Gut.” That’s perfect. Cat gut has long been used to make sutures that will dissolve. The surgeon wants the sutures to hold long enough for the tissue to heal, then the sutures dissolve and disappear. How perfect. May ChatGPT dissolve quickly!

   Thanks for the post, and have a wonderful week! I look forward to the release of DEEP FAKE – another title appropriate for this subject.

   • Cat gut is also a type of guitar string, giving a softer, sound than steel strings… and it’s been replaced these days by nylon – an “artificial catgut” as it were…

     • Thanks for more info on catgut.I didn’t know about guitar strings.

       Artificial is the word of the day.

     • “Catgut” (not made from cats; cattle spinal cord, IIRR) is also used for tennis rackets. I was planning to get my racket restrung with catgut in one direction for spin and nylon in the other for durability. Unfortunately a pinched nerve ended my tennis days.

       • J, the cat lovers of the world celebrate that actual cat guts are not used.

   • Steve, thanks for the medical info on catgut. Unfortunately, I don’t think ChatGPT will dissolve as quickly as stitches.

8. For me, writing is play time, a time to have fun racing through my characters’ stories with them and reporting those stories, including what the characters say and do, “live” as events unfold. No machine can ever replace that, and those who have experienced the sheer joy of creation would ever want it to.

   • “…those who have experienced the sheer joy of creation would ever want it to.”

     You said it, Harvey!

9. So, I know I’m being somewhat more chatty – “intrusive” ? – this morning than usual, but that’s what happens when I check in here while drinking the first cup of Java and then hit ATL’s slog commute… but this appears to me to be much the same thing “production workers” went through with the rise of robotic assembly lines – but, also, to me, it seems that the ones who saw their assembly jobs as more than just a paycheck, found ways to create away from the plant or factory… handmade, custom, unique pieces… which is what writers (and musicians and painter and artists of all stripes), should be about… Artistic Intelligence should always beat out Artificial Intelligence… there’s joy in what we do… something AI will (again, not a present, anyway), ever experience… and that joy is what comes across the page and the ages…

   (I think that’s all… for now… 🙄)

   • George, your participation is always welcome.

     “Artistic Intelligence” is an excellent term for that indefinable human quality of creativity.

10. I think AI will have a limited role in my writing life. I plan to use it like an editor. When I finish my current WIP, I have on my to-do list to use AI to re-write my book blurbs and for Ad taglines. They may be better and they may not. I have two human editors that make my prose better, but they don’t review my blurb or my ads. I don’t see writing an entire novel ever with AI.

    • Alec, using AI as a tool to write blurbs, taglines, and ads makes sense. Many writers (including myself) have trouble distancing ourselves enough from our work to compose objective, effective ad copy.

      Please keep us posted on your results with AI-assisted blurb writing.

      • Debbie, I recommend Dean Wesley Smith’s How to Write Fiction Sales Copy (WMG Writer’s Guides). Copy that into Amazon and you’ll find it, inexpensively, in both paper (recommended) and ebook. My copy is dog-eared.

        • Thanks for the recommendation, Harvey.

11. Good morning, Debbie. AI-“written” faux fiction has been very much on my mind lately. I closely followed what happened at Clarke’s World and read Neil’s tweets and post on it. My friend, author Susan Kaye Quinn has been blogging about it on Facebook. I’ve watched it overtake the art world as well.

    Author Chuck Wendig at his site has a very thoughtful, passionate post on this: https://terribleminds.com/ramble/2023/02/16/a-i-and-the-fetishization-of-ideas/

    Chuck points out that ideas are easy, it’s the implementation of them that is hard, and yet, “ideas” are often considered what is key in writing. Many of us have had the experience of encountering someone who, upon learning we are novelists, offers to provide us with their killer idea, and split the royalties fifty-fifty. He discusses this and more. Well worth the read.

    For my part, the only “A.I.” I will use for my writing is ProWriting Aid or similar. What we call “A.I.” is an algorithm, with no soul, no emotion, no actual thought. It’s programmed to calculate possibility responses in a chat, or project what word or words should go here. It can help with grammar and style checking. Having it write a story is both an ethical minefield and dubious, since creating a truly good story would take a lot of work, just like it does if you do it all by yourself.

    Alan Turning’s test for artificial intelligence was that if a computer hidden behind a current could give responses and comments to a human viewing it on a screen, then it would pass for intelligent. The key thing there is the human interpretation of those responses and comments, and also, our desire for them to be the result of an intelligence, rather than programmed machine learning with no sentience behind it.

    I truly believe that now, more than ever, we must reach deep into our own imaginations, our passions, and bring forth that human creative fire. Each of us has our own unique take on life, on people, our own unique voice, waiting to be rendered in fiction. And, as Jim noted above, our human connection with our readers through that fiction and contact with us in newsletters and elsewhere online, at book signings, etc.

    Thanks for posting about this today! As you can tell, I have some strong feelings about this 🙂

    • Dale, thanks for your insightful comment and the link. Serious discussion among thoughtful authors about AI is encouraging.

      AI is definitely on the zeitgeist today. I just opened an email from Kobo with an article about the ethical use of AI.

12. Thank you, Debbie, for this deep dive into the AI universe.

    I was just thinking what AI software would look like. I imagine databases of names and places. A set of “skeleton” stories for romance novels, mysteries, short stories, etc. Maybe another database of descriptive phrases. It all sounds–well–artificial. It can only provide what it’s been taught to provide, but nothing original.

    As far as writing, why would I give up this self-edifying joy, this creative adventure, for yet another mouse click?

    BTW, I was interested in the statement about plagiarism by Bonifaccic: “The software doesn’t know it’s copying something in violation of an ethical rule that humans apply to themselves.” The software was written by real people who work at a real company. IMO, they are culpable for any plagiarism that’s discovered in their product.

    • Kay, that quote jumped out at me also. AI has no moral compass, no sense of right or wrong, honesty or dishonesty. Like a hammer, it doesn’t differentiate between usefully pounding a nail or brutally bludgeoning a person.

13. My comment: I don’t like it.

    “Nuff said. If I’m grumpy, so be it. 🙂

    • Deb, you have lots of company.

14. Thanks for mentioning my bot job, Debbie. I’ve played around with ChatGPT a lot in the past month, and I’ve recorded a long talk I had with it about fiction writing – approaching it as if I was chatting with the greatest fiction guru the world has ever known.

    It’s really quite amazing what this bot tells you when you learn how to formulate “prompts” so it can give you a meaningful “conclusion”. (Those, I found out, are the operative terms in OpenAI/ChatGPT operation, and the trick is to write the right prompts so it can search for what you want – GIGO.)

    My takeaway in this experiment is that ChatGPT is just another research and writing tool, albeit a very sophisticated tool. It’s not meant to replace human creativity nor should it.

    • It’s not meant to replace human creativity nor should it.

      Whew! Thanks, Garry. I feel better already.

      #changeishard

    • Garry, thanks for doing additional firsthand research. Very interesting. Garbage in garbage out is true in so many aspects of technology.

      Like Deb, I find your conclusion is reassuring.

15. Readers can spot a romance writer who is in it for the money and has no respect for the genre, aka a phony. I doubt an AI can pass that sniff test.

    The platform Quora is now using an AI to generate questions and list “related” articles on the same subject. The questions are nonsense, the related articles are laughably off subject, and they bury the real answers under a sea of dreck.

    So, no, I am not worried that real writers are in any danger of being replaced. But like Quora’s “related” articles, it’s just more dreck that the real books will be buried under at places like Amazon.

    • “…[I]t’s just more dreck that the real books will be buried under at places like Amazon.”

      Aye, there’s the rub. Discoverability is approaching zero.

    • Marilynn, real creativity is buried under mediocrity. Offer a fast, easy shortcut and people will take full advantage of it.

16. The question is whether it would have passed the LSAT if its work had been checked for plagiarism. I doubt it. Plagiarism is not ‘an agreement between humans’. It is illegal and most AI writing at this point is plagiarism.

    • Thanks for addressing plagiarism, JR. A lawyer who hung out his/her shingle after using bots to pass the test is a scary thought.

17. “IT’S ALIVE! IT’S ALIVE!!” –Mary Shelley

    “Mr. Gessler in?” I said.
    He gave me a strange, ingratiating look. “No, sir,” he said, “no. But we can attend to anything with pleasure. We’ve taken the shop over…”
    “Yes, Yes,” I said; “but Mr. Gessler?”
    “Oh!” he answered; “dead.”
    “Dead! But I only received these boots from him last Wednesday week.”
    “Ah!” he said; “a shockin’ go. Poor old man starved ‘imself.”
    “Good God!”
    “Slow starvation, the doctor called it!: ‘ –Galsworthy

    • Actually, in the original novel, as opposed to the screenplay, the monster awakens so:

      “It was already one in the morning; the rain pattered dismally against the panes, and my candle was nearly burnt out, when, by the glimmer of the half-extinguished light, I saw the dull yellow eye of the creature open; it breathed hard, and a convulsive motion agitated its limbs.” –Mary Shelley

    • J, we do seem to be slowing starving our brains with information that has all the nutritional value of fast food.

18. I’ve tinkered with this subject ever since it came up last year in an article in the New Yorker. I tried a free trial with sudowrite and I have tried NovelAI, Dreamily AI and ChatGPT.

    My litmus test for the first three was this: I’d plug in Shelley’s Ozymandias and see what it did with it. The results ranged from pathetic to stupid, in an entertaining way.

    I’ve tried Chatgpt with the plot points from a couple of stories I wrote. One was a story about a trucker who comes in off the road and finds his wife has left him and his doublewide was repossessed. Very dark stuff. ChatGPT produced a sappy feelgood tale with a happy ending and seemed not to know that truckers don’t carry mobile homes with them.

    I do not think it is a substitute for good thoughtful writing, as it is the scrapings of the internet that provide the output. That brings up the possibility that the results you get may have appeared somewhere else before and belong to someone else. Also, what you input may be added to the chatGPT database and used by others without consent.

    ChatGPT is free to use right now but will be monetized soon enough which means that the free ride is going to come to a screeching halt. It’s not cheap to run and the server space and energy usage is a consideration. The financiers behind it are going to want to get paid.

    Of course, aspiring writers are already drowning in a tidal wave of garbage on KDP with people “throwing up something on amazon” and “whee! I’m a published author!” The likelihood is that you’ll get buried in an avalanche of machine made dog waste.

    I taught for three online colleges in the golden age of cut’n’paste plagiarism and I reckon that this chatgpt business will mechanize that process which means that instructors will have to start reading papers before they grade them or edit them. When you see a student’s writing style or quality undergo a shift in the middle of a document it is quite evident what they’re up to, and tools are already available to detect AI written content.

    • I’m hoping that someone will publish an “AI” novel or short chock-full of someone else’s work and get their pants sued off noisily.
      To avoid plagiarism, the program would have to paraphrase everything after generation of a draft file.
      The whole concept is obviously problematic.
      –Ned Ludd

    • “ChatGPT is free to use right now but will be monetized soon enough which means that the free ride is going to come to a screeching halt…The financiers behind it are going to want to get paid.”

      Great point, Robert. This reminds me of the street corner drug dealer who gives out free samples to hook potential customers. Soon those customers are too addicted to escape the trap and the dealer has a nice steady income stream.

19. I don’t doubt that there are readers who like fast, shallow stories pumped out at vomit speed. AI-generated stories would suit them. But there will always be purists who want to discover the writer’s heart within the story and within the story’s imperfections.

    What I wonder about is how small presses weed out AI submissions and how agents keep from getting swamped (maybe they already are) by AI submissions. It can’t be a workable situation; no wonder Clarke’s World closed submissions.

    Maybe publishers and agents need an AI to detect incoming AI!

    • Maybe publishers and agents need an AI to detect incoming AI!

      Then the Frankenscribblers will need another AI to get around the detection AI.

    • Priscilla, getting swamped is one of my concerns for writers submitting their work to publishers. ClarkesWorld obviously felt bad that inevitably good stories are blocked but he couldn’t stay ahead of the deluge of dreck.

20. I tried ChatGPT for a short story that had been languishing on my hard drive for years.

    The scene involves two young women friends. One woman plays a prank on the second making her think she is about to die. When the second finds out it was only a prank, an angry fight breaks out—verbal, not physical.

    I have no experience in what it’s like for two women to get into a vicious argument But I expect it’s a lot different than one between men.

    I asked ChatGPT to write 400 words showing an argument between the women. In less than a minute it turned in the assignment. The result was acceptable but not exceptional. ChatGPT assigned names to the two women. There were words that craft books would tell us are unnecessary and a fair number of -ly adverbs. The dialogue was decent with proper dialogue tags and appropriate action beats.

    I amended the request to ask for more dialogue. It was better. I amended it again to make the second woman more irrational. It added a few new words to meet the request but the piece was largely unchanged. ChatGPT is apparently unaware of how irrational people act. I asked ChatGPT to include profanity but was informed that use of profanity violates OpenAI’s policy. I refined the request a couple more times but each time the changes were more minor.

    My assessment is, ChatGPT did okay but not better than I originally did on my own. It was different but not better. The sentences were technically correct but not engaging. It was blah. I think it would not do well if submitted for a First-page Critique.

    Is it useful? If I were starting a chapter and just couldn’t get that first sentence, I might give the assignment to Chat GPT. It might be easier to edit something to be better than it is to generate a beginning from scratch.

    This entire response was written by me.

    • Chuck, thanks for relating your experience step by step, draft by draft. Bots currently are not good at understanding emotion, intuition, subtlety, and nuance. Will they learn them? I dunno.

      Thanks also for submitting an AI-free response;)

21. If AI stories take over the sales to the majority somehow, then good writers will end up finding wealthy Patrons who are willing to support them to have exclusive access to the ‘good stuff,’ the ‘real stuff.’

    Discerning Patrons. And we will be in Leonardo Da Vinci’s days again.

    • Alicia, that’s a compelling comparison. That may indeed be the future for creative endeavors like literature, art, and music.

      Wanted: a well-heeled angel to support me in the style I’d like to become accustomed to.

22. I went and found one of those AI-written books mentioned in the Business Insider article. It was boring and lifeless. Like, it was telling a story, but it reads like it was trained on the worst passive-voice fanfiction dot net writing you can imagine. It’s technically correct, but … it’s about as exciting as reading the ingredients on a cereal box. Books that are boring won’t sell.

    • Thanks for chiming in, Kessie. That human spark hasn’t yet been duplicated. Stories that lack emotion, humor, irony, nuance, and other subtleties are flat and dull.

      Some cereal boxes may offer better reading!

Comments are closed.