We're in the "wild" early days of this. I hope that some smart techie will come up with a tool that "watermarks" content generated by AI. Readers (buyers) should be aware of what they spend money on, the literary equivalent of an organically grown label. On another note, considering how fewer and fewer people read, flooding a shrinking market with cheap literature doesn't seem to be a money maker... just saying.
Well done, and thank you for this detailed and specific article. I've noticed on recent submissions that I'm asked to vow, essentially, that my work is indeed mine, and that I've not used AI. Since my work - and the work of all those I read and admire - is our way of creating meaning and moving through our own particular days, I'll keep the brain-heart-hand connection going as long as I can. And take a stand against AI "creativity" as long as I have a voice.
I've seen those warnings about AI, but how can they be enforced? I would hope that the readers of submissions are aware enough to see schlock for what it is, but the system doesn't seem fool proof.
Aside from the fact that AI will get better, one problem will be what do you do with credible looking work from a human author who used AI to generate a draft that she or he revised. An increased volume of submissions will also create problems with how unsolicited submissions are reviewed, with new or low credential writers getting short looks or being viewed as suspect.
Thank you for this detailed and alarming article. I'm a fan of Edith Wharton and someone sent me an essay "in the style of" EW to show off what AI-generated writing can do. It was dreadful, had no wit, no color, just sounded vaguely "old-timey."
This clog in the flow of writer produced creativity will need to be approached by the reader with caution for even on-line reviews of a book can be, and often is, plumped with AI air much like frozen turkeys in your supermarket are inflated with air to make them look bigger.
It’s the art of the con pro at its best and worst. New authors will suffer the most as few people will trust that they exist physically.
Several things distress me about the hype around A.I. writing.
First of all, a comment from someone I know in the independent press (i.e. small press) community to the degree that more people are using A.I. in drafting than are willing to openly admit it. This was in the context of a podcast contemplating discussing the WGA and SAG/AFTRA strike, and part of an argument against even discussing the strike. Arguments that A.I. was just one part of the discussion were disregarded and...considering who that person knows, very disturbing.
Second and similar, is the degree to which writers' groups on Facebook end up with so many people defending the use of A.I. in brainstorming, plotting, worldbuilding, and research.
Third is the influx of credibility given this usage by prominent spokespersons such as Joanna Penn, which also turns around and gives organizations she is affiliated with credibility in making assertions about the use of A.I. in writing.
And, finally, the flood of A.I. books on Amazon, including those that imitated Jane Friedman and bogus travel and field guides which have false information.
I just don't get it. I can see plenty of legitimate usages for generative A.I. that don't impinge upon creative endeavors, including agricultural uses where the A.I. crunches, munches, and digests unique field data inputs to help pinpoint soil and plant treatments to conserve the usage of fuel, fertilizer, and other treatments. This focus on creatives strongly suggests that we are being targeted for whatever reasons (the only rationales that make sense to me are outright envy from those who want to have written or created, but not put in the long hours and hard work that goes along with a well-crafted piece of art).
Yes, it does seem odd that AI would appear to be focused so much on areas where it’s not needed (the ability to mass-produce junk). I suspect ChatGPT was as much proof of concept as it was an early glimpse of intended commercial products.
What’s more intriguing to end users and investors alike? A technical description of AI used to identify weather patterns, or an easy-to-use UI that you can feed with text yourself and get immediate results? Investors are human too.
Oh, for sure, ag and weather, often together (rhymes), were probably some of the first AI applications that used the open-source software tools released 6-8 years ago by Google and others. If you look at the bottom of the page for Musixmatch’s AI work, for example, you also see they’re using many of those same tools as well: TensorFlow (Google), PyTorch (Facebook), etc.
Since the amount of arable land is fixed, if not decreasing, much of this agtech work focuses on increasing efficiency and minimizing risk (weather).
But these AI applications don’t capture the attention that the chatbot ones do. To adapt a common farm saying: Nobody talks about agtech with their mouth full.
And yet I find the agtech applications much more fascinating than the chatbots and creative ripoffs. Then again, I've written science fiction that involves a little bit of agtech.
What you're sharing is scary and doesn't bode well for the creative community and it scares the heck out of me. I appreciate your well-researched and well-thought-out article but I would love to read another article that tells writers what we can do about it individually. I love that there are organizations coming out against it - but money speaks, especially in this world, and the bottom line always wins. As far as imitating writers' voices - from what I've heard - AI is designed to "think" and experience the world as humans do. Its imitations of great writers right now may be shoddy, but it will only get better. Are we looking at a future when human creativity will be replaced by a non-human that piggybacks, steals, and replicates what makes us unique?
It is trained to replicate the way humans think and respond and create in the "real" world. So though it's not "in" the world literal world, it is trained learn how we think, what our reactions may be, our emotions, how we create, how we invent, how we express ourselves, how we behave badly and that take that information and both replicate it, change it, invent from it, mix it up. That's what I mean by "in."
The inevitable result of unfettered capitalism twinned with runaway technology. Humans are like lemmings rushing to extinction in search of the latest toy.
A chilling account and thanks for the illumination. A year ago I was asked a question about chatbots and writing to which I cited somebody's opinion (a chatbot's? wouldn't that be ironic) that "it would be used only for the redundant stuff we all have to do every day, boilerplate, not for creative." Boy was I stupid. I've since gotten very worried about the whole thing.
Thanks for the article. If they can, people should also read the referenced Washington Post article. I wonder if AI generated fiction, nonfiction, and poetry or more credible AI assisted work with human revision is already being submitted to lit mags. If that's the case, there will be lots of very bad direct and indirect consequences on literary journals, the market for literary literature, and writing and reading skills generally. I personally also worry if writers, aspirating writers, and people who write to advise writers will see themselves as having no choice but to use AI in, say early drafts, if it becomes the market norm.
I find the alarmist writing in this article to be rather disjointed and full of cliche and non sequitur. Was it written with the help of an AI? I'm joking. (Well, mostly . . . there is a lot of inartfulness to the article, which I normally wouldn't comment on since it's just a substack article, but given the author's explicit mission is to talk up human creativity, it seems fair to make note of it.)
But I do find that the alarmist tone justifies some pretty shaky claims. For example, I highly doubt that AI novels will drive down the price of literature based on the law of supply and demand. As you say, the market for books is already saturated; you can't really get more saturated than it already is. I already have more books on my shelves (and also on my phone! though I wouldn't be surprised if the author is highly suspicious of me reading novels on a non-paper medium) than I will ever be able to read in my lifetime, and so does almost everyone else who reads books. The saturation is in part because more people are publishing, but also because centuries of people have already published. Even without AI, there will always be a mountain of past culture that current authors will have to compete with, and that mountain gets bigger every year. I don't see how flooding the zone with more shitty books is going to alter that dynamic in any significant way.
A similar dynamic had already affected the ability to make a living doing illustration. Even before AI programs like Stable Diffusion and Dall-E came along, the flooding of the marketplace with lots of human talent (aided by the connectivity enabled by the internet) already effectively undercut the ability of most visual artists to make a decent living (here the law of supply really is in effect: much more supply leads to much lower prices). AI may further that trend, but it's an industry that had already been gutted by previous technological and economic trends. Blaming only AI for that is a nice bogeyman story, but it doesn't really comport with reality.
And the baseless claim that Word and GoogleDocs are harvesting the text from your private documents to feed into their AI models really hurts your credibility. The truth is that they don't need those documents to create large language models (and probably wouldn't want them anyway; if you think internet text is often of varying quality, I'd imagine the *average* google docs document to be even lower; most of my google docs are random lists to myself); humanity has already posted trillions of words freely available to anyone with an internet connection. This claim is just baseless fearmongering. I share some of your concerns about AI harming human creative output, but surrounding those concerns with wildly inaccurate claims based on nothing but personal agita against corporations investing in AI only hurts your ability to have your more legitimate claims heard.
Yep, the literary and book markets are oversaturated-- so let's make the situation even worse? Potentially ten times worse? Sorry, I can't buy the logic. Were illustrators having a hard time making a living before AI? Yes-- but many of them were making a living. Now they're outraged. I sat in recently on a zoom session held by a new anti-AI group formed by artists, the Creative Rights Coalition. Yes, believe it, artists are losing jobs and commissions. The same looks to happen to actors and voice actors, screenwriters, musicians-- creatives in a variety of industries and professions. Why the strikes in Hollywood? Mere "agita against corporations"?
As far as what big tech companies are harvesting-- we don't know, yet, the full extent of stolen data, much of it copyrighted. Has ChatGPT4 been fed by LibraryGenesis, for instance, an illegal list of millions of pirated book files? They're not saying. What we do know is tech giants are in a race to perfect the technology, and by all accounts are ruthless. Here's an article, by the way, I've already linked to here, discussing the matter:
Why would they want more data? Why has Twitter/X changed its terms of use, to allow their use of everything we say there?: “you grant us a worldwide, non-exclusive, royalty-free license to use, copy, reproduce, process, adapt, transmit, display” etc. etc. What's the "varying quality" on Twitter?
Am I against people reading non-print books? Well, I have several e-books for sale on Amazon, lost in the flood of content. I also edit an online site. BUT we also do produce print publications, sorry.
Execrable writing has existed forever and, as any reader for a litmag will tell you, is alive and well. So has formulaic writing e.g. penny dreadfuls, Mills and Boon etc. Film and TV writing has ever been thus. Add this to fact that fewer people are reading and buying written works (both paper and digital) tends to suggest that AI generated work is launching into a declining paying demographic that has a shorter and shorter attention span. Original writing has always been a craft industry that rarely produces wealth but has, and will continue to have, a loyal following. Finally, I leave you with a question. If a writer from a third world country with limited English uses AI to build a framework around their unique experiences and delivers a great story for the ages, is that a good or bad thing?
??? Wouldn't they have already written that great story in their own country or own language? Also, presuming the person is brilliant enough to write a great story, and with enough incentive to want the story in English, wouldn't that person have enough incentive and brilliance to learn the English language? See Joseph Conrad, Nabokov, Ayn Rand, and many others who followed that path.
Besides, I'm not talking about individual cases where the devices might have usefulness, but on widespread use of chatbots by hobbyists, opportunists, and content farms alike who'll flood the market with mass production, and have begun doing so. Amid that deluge, how will that "loyal following" you speak of find the good stuff? How will writers crafting truly original work, without the proper connections, funding, and ability to make noise, ever get noticed, by anybody?
(Kind of the dilemma my co-editor and I have with our modest project. We're not sitting on a seven-million dollar endowment. Closer to seven dollars. The market's crowded enough, thank you, without AI!)
Leaving aside the cultural arrogance and ignorance of 3rd world conditions (the 3 writers you refer to were all University educated), there's already a deluge, as I pointed out. Aggregation sites do a great job of pointing the 'true believers' towards the gold amongst the dross and will continue to do so. As the publisher of an even more modest project (Co-editor! Luxury ;-) the search for an audience is perpetual and it was ever thus.
"What about Third World people who might need AI to write?" Such b.s.
Chatbots aren't designed for use by Third World people just trying to survive, who as likely as not are working at starvation wages to support the tech industry.
They're designed for bored bougie people with time on their hands and zero artistic talent because they've spent their lives scamming for more more money.
I can guarantee nobody in the bottom third of American society economically is using a friggin chatbot. They're living in tiny rooms or motels, or trailers, working shitty exhausting jobs, or homeless under freeway overpasses. If they write, it's from dedication to the art of actually writing, often with scribblings in notebooks on work breaks.
Chatbots! The latest tech sales gimmick from hyper-billionaire plutocrats. Meanwhile hundreds of talented writers with stories to tell are shut out because they don't have the proper credentials, the right degree from the just-right Ivy school, or the right cousins in the industry or Uncle Fred, or other connections they just can't play the suck-up kiss-ass literary game. And will never be recognized by your precious alogorithms and data aggregators incapable of understanding human heart and soul.
Our mission at New Pop Lit is to find and present such writers. Here's one of our recent fiction features:
Oh dear, Karl, you really have got your knickers in a twist. All I was suggesting is that no form of IT is purely evil. After all the internet, email etc have provided opportunities undreamed of for disadvantaged people globally. No doubt you and I and all true believers in the art of great writing will continue to recognise it when we see it because, as you say, the human heart and soul cannot be faked. More power to your publishing efforts.
Debate is in my blood, so let me be a contrarian for a moment, please.
1. It doesn't matter what you think about AI. This technology will be as ubiquitous as spellchecker within 5 years. We will learn to live with it regardless of our preferred Luddite leanings. (I have plenty of them; for instance, essentially no Facebook or other social media usage to speak of.)
2. AI will get better at writing. You can't stop that either. It may lack certain creative capabilities for a long time, maybe forever, but the quality of AI-supported writing will dramatically improve with time. Too much money and interest to slow that juggernaut.
3. Some readers may enjoy AI-based stories. So what. Reader tastes, like TV viewing tastes, like all forms of pastimes, aren't my or your concern. Yes, it would be nice to think that our job as writers is to elevate the quality of what is read by most readers, but that's simply not true.
4. Welcome to competition. As writers, we may now face competition from AI-supported stories that get better and better. Our job is to simply be better that them IF we want to continue getting exposure publicly through lit mags and publishers. (This is, of course, entirely optional.)
5. There's precedent in other fields, such as photography, for what to expect. For the past two decades, most photographers developed dual competencies--one, take great pictures with increasingly more sophisticated cameras; two, learn how to improve them through technologies such as Lightroom and Photoshop. For a handful of years, contests sometimes refused to accept digitally-enhanced photos; those days are long gone.
6. Write the way you want, publish the way you want. No one is going to force any of us to use AI, but doing so will not be some cardinal sin of writing any more than using Grammarly or spellchecker, etc. Similarly, thanks to technology, you have vastly more choices for how and where you publish or seek publication. You can create your own Substack or Medium account or submit to a horde of lit mags, most of which would not exist had not technology made them affordable.
The inevitability of AI technology is exactly what Sam Altman and others hopping on the AI bandwagon want everyone to believe. Much of the hype is bluff. Right now AI is an enormous gamble, as all new projects subsisting on speculative VC funding are. We remember the Amazons, but not the hundreds or thousands of other such plans and ideas which vanish into nothing. OpenAI has to make a profit on it first. As we've been seeing, even Twitter has been unable to make a consistent profit, with Elon Musk coming up with more and more ideas of desperation to make it do so.
The internet itself isn't out of the woods yet. (Cliche alert.) It becomes harder and harder even for flagships like The New Yorker to monetize their online content-- and so more and more pop-ups, ads, subscription appeals, heavy discounts. It's not the nirvana for writers and publishers that was promised. I'm also concerned about ever increasing security protocols to sign into accounts-- in some cases you have to take a virtual IQ test or surmount other hurdles just to log on. (How many buses in these photos can you count?) Hackers are becoming ever more proficient.
Our minds trick us into thinking anything that's been around for the course of our lifetime-- even thirty years-- as permanent. I was thinking about this in connection with the old 1939 film Gone With the Wind, which presents a picture of an aristocratic civilization of courtesy and grace which had seemingly been around forever. Study the history and you discover the cotton industry didn't get huge until around 1830-- thirty years before the story takes place-- the mansions and estates-- and the city of Atlanta-- created after that, by freebooting pirates for the most part, who then created a mythology about themselves. Their genteel world interrupted by Northern interlopers: "a civilization gone with the wind." Tech pirates are every bit as much mythmakers.
The point is that nothing in this chaotic world is inevitable or permanent.
With regard to human writers using AI, I'd say there's also an integrity issue if they're submitting to a lit mag or other publisher without disclosing that it'sAI assisted, i.e. not entirely original or generated by the purported author.
ONE WAY TO COMBAT THESE ROBOTS MIGHT BE TO HAVE MORE FLESH AND BLOOD P IUBLIC READINGS WITH NOMINAL ADMISSION COSTS T ATTENDEES SLIDING SCALE. THISISSOMETHING WE CAN CONTROLWITH PREREADING SUBMISSIONS. HAVE THE READINGS,F OR EXAMPLE, AT LOCAL HIGH SCHOOL, COMMUNITY COLLEGE AUDITORIUM AND REGULAR 4 YEAR COLLEGE AUDITORIUMS. WRITERS READ CHAPTERS FROM OUR FICTION BOOKS OR SECTIONS OF POETRY BOOKS.
THIS IS THE TIME TO GET MORE IMAGINATIVE AND FOR WRITERS TO FIND UNTAPPED AUDIENCES.
TO ALL SCHOOLS RE "DIVERSITY AND SOCIAL JUSTICE CONSULTANTS". IN THE 20 YEARS I TAUGHT AT NORTHAMPTON HIGH HSCHOOL, WE HAD A PLETHORA OF PROFESSIONAL DEVELOPMENT DAYS WITH "CONSULTANTS". THEY WERE PAID HUNDREDS OF DOLLARS A DAY FOR PUTTING ON VARIOUS VAPID DOG AND PONY SHOWS THAT WERE USELESS. I CANT REMEMBER A SINGLE WORKSHOP WHERE AN ACTUAL SHORT SSTORY OR POEM WAS READ EXCEPT MAYBE ONE OR TWO TIMES, IN TWENTY YEARS. IT WAS GLARINGLY APPARENT TO ME THAT MOST OF THESE CONSULTANT WERE NOT SERIOUS SCHOLARS OR AFFECIONADOS OF BIOPEC LITERATURE,ART, MUSIC ORFILM. IN ONE WORKSHOOP I ASALKED THEM ABOUT USING IN CLASS, WIT ACCCOMPANYING LITERATURE THE FOLLOWING SOCIAL JUSTICE FILMS WITH ANTIRACIST,ANTISEXIST, PROLABOR , ANTIWAR THEMES: CHARLIECHAPLINS IMMORTAL MODERN TIMES, "I AM AFUGITIVE FROM A CHAIN GANG", "MARKEDWOMAN", "THE GREAT DEBATORS",""STAND AND DELIVER" "JOHNNY GOT HIS GUN","SERGEANT YORK",, "NOTHING BUT A MAN",, DO THE RIGHT THING CHAN IS MISSING. MATEWAM. AND A HOST OFOTHERS. ALSO THERE SOME AMAZING FILMS ABOUT WWII SUCH AS OPEN CITY, PAISAN, THE POLISH WORLD CLASS FILMMAKER ANDREIE WADJA' TRILOGY: A GENERATION,KANAL,A,,NDASHES AND DIAMONDS, ALAIN RESNAI'S "NIGHT AND FOG", "BALLAD OF A SOLDIER" THE CRANES ARE FLYING" KUROSAWA'S BRILLIANT EARLY FILM "NO REGRETS FOR OUR YOUTH", ETTORE SCORZA'S " WE ALL LOVED EACH OTHER SO MUCH", VITTORIO DE SICA'S GENERAL DE ROVERE AND THE BCYCLE THIEF" THE AMERCAN GEM WTH JESSICA LANGAND MAXVON SNYDOW "THE MUSIC BOX", The elegiac "Judgement at Nurenburg", the searing
All the AI hype is not going to stop me for striving to create art with my fiction and poetry. Nor will I let it take away the joy I get from writing.
We're in the "wild" early days of this. I hope that some smart techie will come up with a tool that "watermarks" content generated by AI. Readers (buyers) should be aware of what they spend money on, the literary equivalent of an organically grown label. On another note, considering how fewer and fewer people read, flooding a shrinking market with cheap literature doesn't seem to be a money maker... just saying.
Well done, and thank you for this detailed and specific article. I've noticed on recent submissions that I'm asked to vow, essentially, that my work is indeed mine, and that I've not used AI. Since my work - and the work of all those I read and admire - is our way of creating meaning and moving through our own particular days, I'll keep the brain-heart-hand connection going as long as I can. And take a stand against AI "creativity" as long as I have a voice.
Exactly my thoughts. It’s great to know that we’re a community of artists who have the determination to keep writing in the face of AI schlock.
I've seen those warnings about AI, but how can they be enforced? I would hope that the readers of submissions are aware enough to see schlock for what it is, but the system doesn't seem fool proof.
Aside from the fact that AI will get better, one problem will be what do you do with credible looking work from a human author who used AI to generate a draft that she or he revised. An increased volume of submissions will also create problems with how unsolicited submissions are reviewed, with new or low credential writers getting short looks or being viewed as suspect.
I'll admit I find all this deeply distressing. For those concerned, The Author Guild is inviting signatures for an open letter about this. To process my own fears and to seek a glimmer of hope I wrote a little puece about this which was published by Cleaver. FWIW here it is: https://www.cleavermagazine.com/navigating-back-stargazing-and-the-threat-of-ai-a-craft-essay-by-scott-hurd/
Wonderful comparison with celestial navigation.
Thank you
Thank you for this detailed and alarming article. I'm a fan of Edith Wharton and someone sent me an essay "in the style of" EW to show off what AI-generated writing can do. It was dreadful, had no wit, no color, just sounded vaguely "old-timey."
This tweet from almost exactly a year ago aged well. No one knew what I was talking about just last September: https://twitter.com/timothygreen/status/1573087014001737728
This clog in the flow of writer produced creativity will need to be approached by the reader with caution for even on-line reviews of a book can be, and often is, plumped with AI air much like frozen turkeys in your supermarket are inflated with air to make them look bigger.
It’s the art of the con pro at its best and worst. New authors will suffer the most as few people will trust that they exist physically.
Several things distress me about the hype around A.I. writing.
First of all, a comment from someone I know in the independent press (i.e. small press) community to the degree that more people are using A.I. in drafting than are willing to openly admit it. This was in the context of a podcast contemplating discussing the WGA and SAG/AFTRA strike, and part of an argument against even discussing the strike. Arguments that A.I. was just one part of the discussion were disregarded and...considering who that person knows, very disturbing.
Second and similar, is the degree to which writers' groups on Facebook end up with so many people defending the use of A.I. in brainstorming, plotting, worldbuilding, and research.
Third is the influx of credibility given this usage by prominent spokespersons such as Joanna Penn, which also turns around and gives organizations she is affiliated with credibility in making assertions about the use of A.I. in writing.
And, finally, the flood of A.I. books on Amazon, including those that imitated Jane Friedman and bogus travel and field guides which have false information.
I just don't get it. I can see plenty of legitimate usages for generative A.I. that don't impinge upon creative endeavors, including agricultural uses where the A.I. crunches, munches, and digests unique field data inputs to help pinpoint soil and plant treatments to conserve the usage of fuel, fertilizer, and other treatments. This focus on creatives strongly suggests that we are being targeted for whatever reasons (the only rationales that make sense to me are outright envy from those who want to have written or created, but not put in the long hours and hard work that goes along with a well-crafted piece of art).
Yes, it does seem odd that AI would appear to be focused so much on areas where it’s not needed (the ability to mass-produce junk). I suspect ChatGPT was as much proof of concept as it was an early glimpse of intended commercial products.
What’s more intriguing to end users and investors alike? A technical description of AI used to identify weather patterns, or an easy-to-use UI that you can feed with text yourself and get immediate results? Investors are human too.
That said, there's some significant financial investment in agtech A.I.
Oh, for sure, ag and weather, often together (rhymes), were probably some of the first AI applications that used the open-source software tools released 6-8 years ago by Google and others. If you look at the bottom of the page for Musixmatch’s AI work, for example, you also see they’re using many of those same tools as well: TensorFlow (Google), PyTorch (Facebook), etc.
https://about.musixmatch.com/business/artificial-intelligence
Since the amount of arable land is fixed, if not decreasing, much of this agtech work focuses on increasing efficiency and minimizing risk (weather).
But these AI applications don’t capture the attention that the chatbot ones do. To adapt a common farm saying: Nobody talks about agtech with their mouth full.
And yet I find the agtech applications much more fascinating than the chatbots and creative ripoffs. Then again, I've written science fiction that involves a little bit of agtech.
What you're sharing is scary and doesn't bode well for the creative community and it scares the heck out of me. I appreciate your well-researched and well-thought-out article but I would love to read another article that tells writers what we can do about it individually. I love that there are organizations coming out against it - but money speaks, especially in this world, and the bottom line always wins. As far as imitating writers' voices - from what I've heard - AI is designed to "think" and experience the world as humans do. Its imitations of great writers right now may be shoddy, but it will only get better. Are we looking at a future when human creativity will be replaced by a non-human that piggybacks, steals, and replicates what makes us unique?
How can it experience the world when it is not *in* the world?
It is trained to replicate the way humans think and respond and create in the "real" world. So though it's not "in" the world literal world, it is trained learn how we think, what our reactions may be, our emotions, how we create, how we invent, how we express ourselves, how we behave badly and that take that information and both replicate it, change it, invent from it, mix it up. That's what I mean by "in."
That doesn't sound like invention or creation. It is a fancy parrot. https://www.forbes.com/sites/bernardmarr/2023/03/27/the-intersection-of-ai-and-human-creativity-can-machines-really-be-creative/?sh=5387c383dbc4
The inevitable result of unfettered capitalism twinned with runaway technology. Humans are like lemmings rushing to extinction in search of the latest toy.
A chilling account and thanks for the illumination. A year ago I was asked a question about chatbots and writing to which I cited somebody's opinion (a chatbot's? wouldn't that be ironic) that "it would be used only for the redundant stuff we all have to do every day, boilerplate, not for creative." Boy was I stupid. I've since gotten very worried about the whole thing.
Thanks for the article. If they can, people should also read the referenced Washington Post article. I wonder if AI generated fiction, nonfiction, and poetry or more credible AI assisted work with human revision is already being submitted to lit mags. If that's the case, there will be lots of very bad direct and indirect consequences on literary journals, the market for literary literature, and writing and reading skills generally. I personally also worry if writers, aspirating writers, and people who write to advise writers will see themselves as having no choice but to use AI in, say early drafts, if it becomes the market norm.
I find the alarmist writing in this article to be rather disjointed and full of cliche and non sequitur. Was it written with the help of an AI? I'm joking. (Well, mostly . . . there is a lot of inartfulness to the article, which I normally wouldn't comment on since it's just a substack article, but given the author's explicit mission is to talk up human creativity, it seems fair to make note of it.)
But I do find that the alarmist tone justifies some pretty shaky claims. For example, I highly doubt that AI novels will drive down the price of literature based on the law of supply and demand. As you say, the market for books is already saturated; you can't really get more saturated than it already is. I already have more books on my shelves (and also on my phone! though I wouldn't be surprised if the author is highly suspicious of me reading novels on a non-paper medium) than I will ever be able to read in my lifetime, and so does almost everyone else who reads books. The saturation is in part because more people are publishing, but also because centuries of people have already published. Even without AI, there will always be a mountain of past culture that current authors will have to compete with, and that mountain gets bigger every year. I don't see how flooding the zone with more shitty books is going to alter that dynamic in any significant way.
A similar dynamic had already affected the ability to make a living doing illustration. Even before AI programs like Stable Diffusion and Dall-E came along, the flooding of the marketplace with lots of human talent (aided by the connectivity enabled by the internet) already effectively undercut the ability of most visual artists to make a decent living (here the law of supply really is in effect: much more supply leads to much lower prices). AI may further that trend, but it's an industry that had already been gutted by previous technological and economic trends. Blaming only AI for that is a nice bogeyman story, but it doesn't really comport with reality.
And the baseless claim that Word and GoogleDocs are harvesting the text from your private documents to feed into their AI models really hurts your credibility. The truth is that they don't need those documents to create large language models (and probably wouldn't want them anyway; if you think internet text is often of varying quality, I'd imagine the *average* google docs document to be even lower; most of my google docs are random lists to myself); humanity has already posted trillions of words freely available to anyone with an internet connection. This claim is just baseless fearmongering. I share some of your concerns about AI harming human creative output, but surrounding those concerns with wildly inaccurate claims based on nothing but personal agita against corporations investing in AI only hurts your ability to have your more legitimate claims heard.
Yep, the literary and book markets are oversaturated-- so let's make the situation even worse? Potentially ten times worse? Sorry, I can't buy the logic. Were illustrators having a hard time making a living before AI? Yes-- but many of them were making a living. Now they're outraged. I sat in recently on a zoom session held by a new anti-AI group formed by artists, the Creative Rights Coalition. Yes, believe it, artists are losing jobs and commissions. The same looks to happen to actors and voice actors, screenwriters, musicians-- creatives in a variety of industries and professions. Why the strikes in Hollywood? Mere "agita against corporations"?
As far as what big tech companies are harvesting-- we don't know, yet, the full extent of stolen data, much of it copyrighted. Has ChatGPT4 been fed by LibraryGenesis, for instance, an illegal list of millions of pirated book files? They're not saying. What we do know is tech giants are in a race to perfect the technology, and by all accounts are ruthless. Here's an article, by the way, I've already linked to here, discussing the matter:
https://www.laptopmag.com/features/google-docs-ai-is-it-safe-im-a-novel-writer-and-tech-journalist-lets-talk
Why would they want more data? Why has Twitter/X changed its terms of use, to allow their use of everything we say there?: “you grant us a worldwide, non-exclusive, royalty-free license to use, copy, reproduce, process, adapt, transmit, display” etc. etc. What's the "varying quality" on Twitter?
Am I against people reading non-print books? Well, I have several e-books for sale on Amazon, lost in the flood of content. I also edit an online site. BUT we also do produce print publications, sorry.
And here we have it, folx. The standard disclaimer by an A.I. supporter.
Execrable writing has existed forever and, as any reader for a litmag will tell you, is alive and well. So has formulaic writing e.g. penny dreadfuls, Mills and Boon etc. Film and TV writing has ever been thus. Add this to fact that fewer people are reading and buying written works (both paper and digital) tends to suggest that AI generated work is launching into a declining paying demographic that has a shorter and shorter attention span. Original writing has always been a craft industry that rarely produces wealth but has, and will continue to have, a loyal following. Finally, I leave you with a question. If a writer from a third world country with limited English uses AI to build a framework around their unique experiences and delivers a great story for the ages, is that a good or bad thing?
??? Wouldn't they have already written that great story in their own country or own language? Also, presuming the person is brilliant enough to write a great story, and with enough incentive to want the story in English, wouldn't that person have enough incentive and brilliance to learn the English language? See Joseph Conrad, Nabokov, Ayn Rand, and many others who followed that path.
Besides, I'm not talking about individual cases where the devices might have usefulness, but on widespread use of chatbots by hobbyists, opportunists, and content farms alike who'll flood the market with mass production, and have begun doing so. Amid that deluge, how will that "loyal following" you speak of find the good stuff? How will writers crafting truly original work, without the proper connections, funding, and ability to make noise, ever get noticed, by anybody?
(Kind of the dilemma my co-editor and I have with our modest project. We're not sitting on a seven-million dollar endowment. Closer to seven dollars. The market's crowded enough, thank you, without AI!)
Leaving aside the cultural arrogance and ignorance of 3rd world conditions (the 3 writers you refer to were all University educated), there's already a deluge, as I pointed out. Aggregation sites do a great job of pointing the 'true believers' towards the gold amongst the dross and will continue to do so. As the publisher of an even more modest project (Co-editor! Luxury ;-) the search for an audience is perpetual and it was ever thus.
"What about Third World people who might need AI to write?" Such b.s.
Chatbots aren't designed for use by Third World people just trying to survive, who as likely as not are working at starvation wages to support the tech industry.
(See: https://time.com/6247678/openai-chatgpt-kenya-workers/)
They're designed for bored bougie people with time on their hands and zero artistic talent because they've spent their lives scamming for more more money.
I can guarantee nobody in the bottom third of American society economically is using a friggin chatbot. They're living in tiny rooms or motels, or trailers, working shitty exhausting jobs, or homeless under freeway overpasses. If they write, it's from dedication to the art of actually writing, often with scribblings in notebooks on work breaks.
Chatbots! The latest tech sales gimmick from hyper-billionaire plutocrats. Meanwhile hundreds of talented writers with stories to tell are shut out because they don't have the proper credentials, the right degree from the just-right Ivy school, or the right cousins in the industry or Uncle Fred, or other connections they just can't play the suck-up kiss-ass literary game. And will never be recognized by your precious alogorithms and data aggregators incapable of understanding human heart and soul.
Our mission at New Pop Lit is to find and present such writers. Here's one of our recent fiction features:
https://newpoplit.com/portfolio/tales-along-turtle-heart-road/
Here's another:
https://newpoplit.com/portfolio/glow-worm-farm/
Compassion, humanity, and understanding which will never come from a soulless device.
Oh dear, Karl, you really have got your knickers in a twist. All I was suggesting is that no form of IT is purely evil. After all the internet, email etc have provided opportunities undreamed of for disadvantaged people globally. No doubt you and I and all true believers in the art of great writing will continue to recognise it when we see it because, as you say, the human heart and soul cannot be faked. More power to your publishing efforts.
Debate is in my blood, so let me be a contrarian for a moment, please.
1. It doesn't matter what you think about AI. This technology will be as ubiquitous as spellchecker within 5 years. We will learn to live with it regardless of our preferred Luddite leanings. (I have plenty of them; for instance, essentially no Facebook or other social media usage to speak of.)
2. AI will get better at writing. You can't stop that either. It may lack certain creative capabilities for a long time, maybe forever, but the quality of AI-supported writing will dramatically improve with time. Too much money and interest to slow that juggernaut.
3. Some readers may enjoy AI-based stories. So what. Reader tastes, like TV viewing tastes, like all forms of pastimes, aren't my or your concern. Yes, it would be nice to think that our job as writers is to elevate the quality of what is read by most readers, but that's simply not true.
4. Welcome to competition. As writers, we may now face competition from AI-supported stories that get better and better. Our job is to simply be better that them IF we want to continue getting exposure publicly through lit mags and publishers. (This is, of course, entirely optional.)
5. There's precedent in other fields, such as photography, for what to expect. For the past two decades, most photographers developed dual competencies--one, take great pictures with increasingly more sophisticated cameras; two, learn how to improve them through technologies such as Lightroom and Photoshop. For a handful of years, contests sometimes refused to accept digitally-enhanced photos; those days are long gone.
6. Write the way you want, publish the way you want. No one is going to force any of us to use AI, but doing so will not be some cardinal sin of writing any more than using Grammarly or spellchecker, etc. Similarly, thanks to technology, you have vastly more choices for how and where you publish or seek publication. You can create your own Substack or Medium account or submit to a horde of lit mags, most of which would not exist had not technology made them affordable.
7. Welcome to the writer's brave new world.
The inevitability of AI technology is exactly what Sam Altman and others hopping on the AI bandwagon want everyone to believe. Much of the hype is bluff. Right now AI is an enormous gamble, as all new projects subsisting on speculative VC funding are. We remember the Amazons, but not the hundreds or thousands of other such plans and ideas which vanish into nothing. OpenAI has to make a profit on it first. As we've been seeing, even Twitter has been unable to make a consistent profit, with Elon Musk coming up with more and more ideas of desperation to make it do so.
The internet itself isn't out of the woods yet. (Cliche alert.) It becomes harder and harder even for flagships like The New Yorker to monetize their online content-- and so more and more pop-ups, ads, subscription appeals, heavy discounts. It's not the nirvana for writers and publishers that was promised. I'm also concerned about ever increasing security protocols to sign into accounts-- in some cases you have to take a virtual IQ test or surmount other hurdles just to log on. (How many buses in these photos can you count?) Hackers are becoming ever more proficient.
Our minds trick us into thinking anything that's been around for the course of our lifetime-- even thirty years-- as permanent. I was thinking about this in connection with the old 1939 film Gone With the Wind, which presents a picture of an aristocratic civilization of courtesy and grace which had seemingly been around forever. Study the history and you discover the cotton industry didn't get huge until around 1830-- thirty years before the story takes place-- the mansions and estates-- and the city of Atlanta-- created after that, by freebooting pirates for the most part, who then created a mythology about themselves. Their genteel world interrupted by Northern interlopers: "a civilization gone with the wind." Tech pirates are every bit as much mythmakers.
The point is that nothing in this chaotic world is inevitable or permanent.
And another A.I. defender. Sigh.
With regard to human writers using AI, I'd say there's also an integrity issue if they're submitting to a lit mag or other publisher without disclosing that it'sAI assisted, i.e. not entirely original or generated by the purported author.
It sounds the use of AI for drafts by human writers is already being normalized.
ONE WAY TO COMBAT THESE ROBOTS MIGHT BE TO HAVE MORE FLESH AND BLOOD P IUBLIC READINGS WITH NOMINAL ADMISSION COSTS T ATTENDEES SLIDING SCALE. THISISSOMETHING WE CAN CONTROLWITH PREREADING SUBMISSIONS. HAVE THE READINGS,F OR EXAMPLE, AT LOCAL HIGH SCHOOL, COMMUNITY COLLEGE AUDITORIUM AND REGULAR 4 YEAR COLLEGE AUDITORIUMS. WRITERS READ CHAPTERS FROM OUR FICTION BOOKS OR SECTIONS OF POETRY BOOKS.
THIS IS THE TIME TO GET MORE IMAGINATIVE AND FOR WRITERS TO FIND UNTAPPED AUDIENCES.
TO ALL SCHOOLS RE "DIVERSITY AND SOCIAL JUSTICE CONSULTANTS". IN THE 20 YEARS I TAUGHT AT NORTHAMPTON HIGH HSCHOOL, WE HAD A PLETHORA OF PROFESSIONAL DEVELOPMENT DAYS WITH "CONSULTANTS". THEY WERE PAID HUNDREDS OF DOLLARS A DAY FOR PUTTING ON VARIOUS VAPID DOG AND PONY SHOWS THAT WERE USELESS. I CANT REMEMBER A SINGLE WORKSHOP WHERE AN ACTUAL SHORT SSTORY OR POEM WAS READ EXCEPT MAYBE ONE OR TWO TIMES, IN TWENTY YEARS. IT WAS GLARINGLY APPARENT TO ME THAT MOST OF THESE CONSULTANT WERE NOT SERIOUS SCHOLARS OR AFFECIONADOS OF BIOPEC LITERATURE,ART, MUSIC ORFILM. IN ONE WORKSHOOP I ASALKED THEM ABOUT USING IN CLASS, WIT ACCCOMPANYING LITERATURE THE FOLLOWING SOCIAL JUSTICE FILMS WITH ANTIRACIST,ANTISEXIST, PROLABOR , ANTIWAR THEMES: CHARLIECHAPLINS IMMORTAL MODERN TIMES, "I AM AFUGITIVE FROM A CHAIN GANG", "MARKEDWOMAN", "THE GREAT DEBATORS",""STAND AND DELIVER" "JOHNNY GOT HIS GUN","SERGEANT YORK",, "NOTHING BUT A MAN",, DO THE RIGHT THING CHAN IS MISSING. MATEWAM. AND A HOST OFOTHERS. ALSO THERE SOME AMAZING FILMS ABOUT WWII SUCH AS OPEN CITY, PAISAN, THE POLISH WORLD CLASS FILMMAKER ANDREIE WADJA' TRILOGY: A GENERATION,KANAL,A,,NDASHES AND DIAMONDS, ALAIN RESNAI'S "NIGHT AND FOG", "BALLAD OF A SOLDIER" THE CRANES ARE FLYING" KUROSAWA'S BRILLIANT EARLY FILM "NO REGRETS FOR OUR YOUTH", ETTORE SCORZA'S " WE ALL LOVED EACH OTHER SO MUCH", VITTORIO DE SICA'S GENERAL DE ROVERE AND THE BCYCLE THIEF" THE AMERCAN GEM WTH JESSICA LANGAND MAXVON SNYDOW "THE MUSIC BOX", The elegiac "Judgement at Nurenburg", the searing
"
?