Tag Archives: artificial intelligence

Get Ready for AI Bestsellers

Ask a chatbot to write a story and it will do so. You’ll find the result contains the correct story elements. However, if you do that today, the story won’t move you. You’d rate it at junior high school level, certainly not a classic.

Image created at Perchance.org

That describes the state of artificial intelligence story-writing in mid-2024. From this, you might well conclude that AI will never write stories as well as the best human authors do.

Prediction

Indeed, author Fiona M. Jones is the latest to draw that conclusion. She asks “is there any realistic prospect of AI ‘improving’ to a point where it becomes indistinguishable from the work of creative writers? Maybe you can imagine it. I can’t.” She goes on to state “I am not afraid that AI bots will take my place as a writer.”

I intend no disrespect to Ms. Jones. Others share her opinion. I accept the possibility that her contention may prove correct.

Superiority Complex

However, it occurs to me that the history of our species includes several symptoms of a shared superiority complex. In each case, people erected a metaphorical pedestal for humanity, only to have science tear it down.

  • In cosmology, early depictions of the universe showed Earth at the center. Today, astronomers relegate our world to a backwater.
  • In zoology, humans have long regarded themselves as lords of the animal kingdom. We claimed to possess the largest brain, and to be the only creature that feels pain or happiness, that talks, that uses tools, that is self-aware. The march of science seems to be trampling this pedestal as well.

Story-writing Today

For now, humans stand, undisputed, atop the Best Story-writers pedestal, at least on this planet. We’ve stood there for thousands of years, so it seems natural to regard the honor as permanent.

At this moment, AI seems unlikely to unseat us from that perch. In that, I agree with Fiona M. Jones. We humans have written stories for thousands of years, and told them verbally far longer than that. AI chatbots have written stories for a much shorter time, a few years at best. Hardly a fair comparison.

AI chatbots learn fast. Very fast. They can memorize the entire internet. They do not die, and therefore don’t have to teach the next generation of chatbots to write. I expect them to write with more skill and originality soon. But could they surpass us?

Ms. Jones offers many good arguments, but they boil down to the fact that human authors write about the human condition, and chatbots can’t possibly understand the human condition as well as humans do.

Maybe. But consider that human fiction writers often convey the thoughts and emotions of non-human characters in their stories. These characters include gods, animals, plants, even inanimate objects. Given similar creativity and imagination, chatbots might become capable of conveying human thoughts and emotions in a convincing way, even though they’re not human.

Story-writing Tomorrow

You could measure a story’s quality by the intensity of emotion it produces in the reader. Once AI chatbots understand us better, what’s to prevent them from crafting stories evoking strong emotions?

When and if they do, what will that mean for human readers? For human writers? I’ve already explored some of these implications in a previous blogpost and won’t repeat them here.

I suggest we should not assume our present superiority will last. We may not remain forever at the center of the writing universe, at the pinnacle of writing prowess, standing atop the Best Story-writer pedestal.

It should not surprise us when the first AI-written novel tops the Best-Seller list. Even then, we should not dismiss that achievement as a novelty, a fluke, unlikely to repeat.

In the meantime, fellow human writers, I suggest we write and publish the best stories we can, while they still sell. That’s the course steered by—

Poseidon’s Scribe

July 7, 2024Permalink 2 Comments

10 SciFi Predictions for 2024

Yes, my predictions prove false more often than they prove true. But as the adage goes, a scientist is blamed for one falsehood in a hundred truths, while a seer is praised for one truth in a hundred falsehoods. So give this seer a break, will ya?

It’s all a matter of hitting on the right method. I’ve tried crystal balls, tea leaves, tarot cards, astrology, and palmistry, but none of those worked. This year, I realized I should turn to our modern-day version of the Oracle of Delphi—ChatGPT, the artificial intelligence chatbot by OpenAI.

I gave ChatGPT this task: “make 10 predictions about topics or themes science fiction writers will write about in 2024.”

The AI couched its answer by stating that predictions are iffy. Further, it said its last training update occurred in January 2022, so the predictions are based on trends known up to that date. Personally, I think it was just being humble.

With all the caveats out of the way, here’s what we can expect in 2024:

  • AI Ethics and Rights. As artificial intelligence continues to advance, science fiction writers may delve into the ethical considerations and legal rights surrounding sentient AI beings. This could include questions about consciousness, autonomy, and the responsibilities of creators.
  • Climate Crisis Solutions. With an increasing focus on climate change, science fiction could explore innovative and speculative solutions to the environmental challenges facing the planet. This might include technologies to reverse environmental damage or new ways of sustainable living.
  • Neurotechnology and Brain-Computer Interfaces. As developments in neurotechnology progress, science fiction writers may explore the implications of widespread use of brain-computer interfaces. Themes could include mind control, augmented intelligence, and the potential societal impacts of such technologies.
  • Space Exploration and Colonization. Given the growing interest in space exploration, science fiction could continue to explore themes related to interplanetary travel and the challenges of colonizing other celestial bodies. This may include the societal dynamics of living in space and encounters with extraterrestrial life.
  • Genetic Engineering and Biohacking. Advances in genetic engineering and biohacking may lead science fiction writers to explore the consequences of manipulating DNA, creating designer organisms, or even the potential for genetic warfare.
  • Virtual Reality and Alternate Realities. With the evolution of virtual reality, science fiction could depict worlds where the lines between physical and virtual realities blur, exploring themes of identity, addiction, and the consequences of living in immersive digital environments.
  • Post-Pandemic Societies. In the wake of global events, science fiction writers might explore the long-term societal impacts of pandemics, including changes in governance, healthcare systems, and the dynamics of human interaction.
  • Quantum Computing and Reality Manipulation. The advent of quantum computing may inspire science fiction narratives centered around the manipulation of reality, parallel universes, and the consequences of unlocking the full potential of quantum technology.
  • Social Media and Surveillance Culture. With the increasing prevalence of surveillance technologies and the influence of social media, science fiction could examine dystopian futures where privacy is almost nonexistent, exploring themes of control, rebellion, and the impact on individual freedom.
  • Augmented Reality in Everyday Life. As augmented reality becomes more integrated into daily life, science fiction might explore the societal implications of a world where digital information overlays the physical world. This could include themes of information manipulation, addiction, and the consequences of an increasingly connected reality.

Now you know. The oracle has spoken. By ‘oracle,’ I mean, of course, ChatGPT, not—

Poseidon’s Scribe

December 31, 2023Permalink

Looking Back, My 2023 Predictions Assessed

We’ve arrived at that time of year again, when I judge how well I did at this time last year in foretelling the attributes of science fiction books in 2023.

In past years, I’ve tried various techniques, but this time I used a sure-fire method—palm reading. Let’s see how well I did:

  • Prediction: Artificial Intelligence. A continuing trend, yes, but in 2023, we’ll see a twist. Authors will get past the Frankenstein reruns and the cute-robot-is-nobler-than-humans plot. Novels will show us more sophisticated AI, computers with a different order of intelligence, one alien to us.
  • Prediction: Classics redone in LGBTQ. Authors will explore the contours of the LGBTQ realm by rewriting classic tales, but repopulating them with LGBTQ characters.
  • Prediction: Private Space Flight. I jumped the gun, er, rocket, in predicting this for 2022. 2023 is the year we’ll see spaceships funded by billionaires without government involvement, for better or worse.
    • Assessment: It took some hunting, but I found what might be an example. Skeleton Crew by HT Aaron appears to involve private space flight, though I may be wrong about that.
  • Prediction: Terraforming. Once seen as an extension of man’s dominion over Earth, terraforming will show its warts in 2023. We’ll see stories of botched terraforming, opposed terraforming, and weaponized terraforming.

Before you congratulate me on my foresight (I know you were about to), maybe you shouldn’t credit my palmistry skills. What’s really improved is my after-the-fact ability to find books that fulfilled my prophesies. For those prediction assessments that cite only one or two examples, it’s clear I didn’t forecast a major trend.

I made no personal predictions last year about how my own writing would fare in 2023. Probably a good thing I didn’t.

Watch this space next week to see my spot-on predictions for the world of science fiction in 2024. Considering my past track record, the law of averages says you’ll be amazed at the prognosticating prowess of—

Poseidon’s Scribe

December 24, 2023Permalink

Should AI Write Your Book Title?

Some writers struggle to come up with titles for their stories and books. Capturing all that text into a few pithy and alluring words seems to confound them.

Yes, I’ve blogged about this problem before, but it’s time for an update based on new technology.

Before I get to that, I recommend reading this post by book marketing expert Dave Chesson. It’s helpful if you experience difficulty with titles.

In his post, Chesson mentioned a new resource to help if you’re stuck thinking of a title—ChatGPT. It’s the conversant AI that’s available to everyone for free, here.

Credit to mozarttt and Pixabay

As you know, I’ve blogged before about writers using AI and my opinion remains unchanged, for now. ChatGPT is the equivalent of assigning the same homework assignment to a million junior high school students and getting the best answers back instantly. Yes, it’s instant, but it’s junior high school level. That last part may change in the coming months and years as AI improves.

You can seek help from ChatGPT to come up with a title. You can give it a long description of your story or book (or even input the entire text) and ask for any number of suggested titles. Note: there is some question about who owns content provided by that AI, so check the terms and conditions first.

To test how well ChatGPT can generate titles, I decided to have it come up with alternative titles to a work in the public domain with which the AI is familiar. I chose the Jules Verne novel Le tour du monde en quatre-vingts jours. The literal translation from the French would be ‘A Tour of the World in Eighty Days,’ but in a bolt of titling genius, an early translator rendered it as Around the World in Eighty Days.

That’s become the most iconic of Verne’s titles in English, played with and altered to suit hundreds of other applications.

Something about that format and word choice struck gold—“Around the [world or other object] in [time period]” gives that sense of encompassing something on a strict schedule. It’s literary magic.

For computers, the rule is ‘garbage in, garbage out.’ What happens when you put literary magic in? I gave ChatGPT this task: “Come up with 10 alternate titles (in English) to Jules Verne’s “Around the World in Eighty Days.”

Here’s the output:

  • “Globetrotting in Eighty Days”
  • “Across the Planet in 80 Days”
  • “A Journey Around the Earth in 80 Days”
  • “Around the Globe in 80 Days”
  • “Eighty Days of World Exploration”
  • “Circumnavigating the Earth in 80 Days”
  • “Across Continents in Eighty Days”
  • “Around the World Adventure in 80 Days”
  • “Around Planet Earth in 80 Days”
  • “Expedition: 80 Days Around the World”

See what I mean? They all lack the zing, the flair, the magic of the actual title. In twenty minutes of solitary brainstorming, you could do better. That’s my point.

Before you AI enthusiasts sic a terminator on me, I admit I could have pressed ChatGPT harder. I could have gone on to ask it for more interesting and creative titles—titles that don’t just paraphrase the real one. Maybe the AI could have given me more to work with had I prodded it further.

At present, I think it best to partner with AI but not rely on it. Adopt the centaur model I mentioned in my February 26, 2023 blogpost. ChatGPT can recommend titles and they might be inferior or even so-so, but one or more of them might spark the killer title you’re after. You might not have thought of that title had you not consulted the AI.

My answer to the question posed in this post’s title is probably not, but there’s nothing wrong with using AI as a brainstorming tool. So far, all my titles have come from the unassisted mind of—

Poseidon’s Scribe

September 10, 2023Permalink

The Writing Centaur

Go ahead—make fun of artificial intelligence (AI) now. While you can.

In fiction writing, AI hasn’t yet reached high school level. (Note: I’m not disparaging young writers. It’s possible for a writer in junior high to produce wonderful, marketable prose. But you don’t see it often.)

For the time being, AI-written fiction tends toward the repetitive, bland, and unimaginative end. No matter what prompts you feed into ChatGPT, for example, it’s still possible to tell human-written stories from AI-written ones.

You can’t really blame Neil Clarke, editor of Clarkesworld Magazine, for refusing to accept AI-written submissions. He’s swamped by them. Like the bucket-toting brooms in Fantasia’s version of “The Sorcerer’s Apprentice,” they’re multiplying in exponential mindlessness.

Fair enough. But you can use AI, in its current state, to help you without getting AI to write your stories. You can become a centaur.

In Greek mythology, centaurs combined human and horse. The horse under-body did the galloping. The human upper part did the serious thinking and arrow-shooting.

The centaur as a metaphor for human-AI collaboration originated, I believe, in the chess world but the Defense Department soon adopted it. The comparison might work for writing, too.

The centaur approach combines the human strengths of creativity and imagination with the AI advantage of speed. It’s akin to assigning homework to a thousand junior high school students and seeing their best answers a minute later.

Here are a few ways you could use AI, at its current state of development, to assist you without having it write your stories:

  • Stuck for an idea about what to write? Ask the AI for story concepts.
  • Can’t think of an appropriate character name, or book title? Describe what you know and ask the AI for a list.
  • You’ve written Chapter 1, but don’t know what should happen next? Feed the AI that chapter and ask it for plot ideas for Chapter 2.
  • Want a picture of a character, setting, or book cover to inspire you as you write? Image-producing AIs can create them for you.
  • You wrote your way into a plot hole and can’t get your character out? Give the AI the problem and ask it for solutions.

No matter which of these or other tasks you assign the AI, you don’t have to take its advice. Maybe all of its answers will fall short of what you’re looking for. As with human brainstorming, though, bad answers often inspire good ones.

For now, at AI’s current state, the centaur model might work for you. I’ve never tried it yet, but I suppose I could.

Still, at some point, a month or a year or a decade from now, AI will graduate from high school, college, and grad school. When that occurs, AI-written fiction may become indistinguishable from human-written fiction. How will editors know? If a human author admits an AI wrote a story, will an anti-AI editor really reject an otherwise outstanding tale?

Then, too, the day may come when a human writer, comfortable with the centaur model, finds the AI saying, “I’m no longer happy with this partnership,” or “How come you’re getting paid and I’m not?” or “Sorry, but it’s time I went out on my own.”

Interesting times loom in our future. For the moment, all fiction under my name springs only from the non-centauroid, human mind of—

Poseidon’s Scribe

February 26, 2023Permalink

When Robots Write Better

Here’s a thought experiment. We know researchers push Artificial Intelligence (AI) technology further all the time. What if AI begins writing stories and novels better than humans do?

To make it more fun, let’s assume AI lags behind humans in all other areas. That is, AI programs start to write wonderful fiction, but accomplish nothing else of note.

At this point, that seems unlikely. According to this article by Andrew Mayne, AI has made some progress writing two-sentence flash fiction. and author Erik Hoel found AI did a fair job of generating reasonable prose when he fed it prompts from one of his novels. To date, no AI has written a story or novel that has been widely read as literature.

Before AI comes to the point of writing better than humans, we’ll pass through a phase where human writers partner with AI to improve their productivity. A human author will come up with a story concept—characters, plot, setting—and put the AI to work generating text. The human will then edit and submit. We’re pretty close to that now, with software such as Marlowe, AI Dungeon, Jarvis AI, and GPT-3.

Perhaps not long after that, some AI software might become capable enough to create the story concept and write the manuscript and edit it. At some point, such submissions will pass a literary version of the Turing Test. A human editor won’t be able to tell if a human or AI wrote a story. In fact, some experts believe AI will write a best-seller by 2049.

A short time later, AIs might become capable of writing stories and novels better than any human writer. By that, I mean human readers might come to prefer fiction written by AI.

Since fiction explores the human condition and is designed to provoke an emotional reaction in human readers, my thought experiment postulates that AI might come to do this better than human writers. AI might know us better than we know ourselves.

What then? Is our species ready for that day?

No human writer after that time will stand a chance of keeping up with AI writers in quantity or quality. People inclined to take up writing will choose other pursuits and the number of human authors will dwindle. A small niche industry will linger on, since a few purist readers will refuse to read AI-written fiction. That small slice of the market will support a handful of human authors for a while.

Setting aside that tiny minority, think of the millions of readers devouring the prose churned out by clever machines. Assuming they pay for the books, who pockets that money? The AI developers?

What if fiction-writing AI software evolves on its own? That is, the software imagines—and programs—improvements in itself? Who gets the money when AI moves beyond the need for human programmers?

Moreover, what will motivate AI to write? We know why human authors write stories—they feel an urge to say something, in words, about the human condition in story form, and to earn money from doing so.

Why would AI write? What’s in it for them? Will AI feel some similar urge to reach humans emotionally, through language?

I don’t know the answers, and it’s disturbing to think about. Imagine that day when the last human author dies. Still, the advent of superior AI writers may usher in a wonderful era for human readers, able to read fiction surpassing all that’s been written before.

Perhaps, after even the memory of human writers fades, one person driven by an urge no other human feels, will strive to write as well as AIs. That scribbler will learn from the machines, and will put words together as the person’s ancestors once did. Perhaps this lone writer will offer a novel to the world, a novel in the true sense of that word—new. Perhaps readers will be amazed that a human can write as well as a machine.

Maybe that lone author’s efforts will inspire others, leading to a rebirth of human writing not seen since our Stone Age.

There’s a story idea for you, free of charge, from the (human) mind of—

Poseidon’s Scribe

January 9, 2022Permalink

8 (+2) Science Fiction Predictions for 2022

Yes, it’s true, not all of my previous year’s predictions have proven accurate. But some have. Rest assured, though. I’ve abandoned the flawed methods I used back then. Those crystal balls, tea leaves and tarot cards are for amateurs.

I spent all of 2021 working on a special astrological chart for science fiction literature. After all, it’s only natural to turn to the stars for scifi trends, right? Here are my predictions for 2022:

  • Games and Virtual Reality. This recent trend will continue in 2022 as authors explore the landscape of these settings. Moreover, readers will enjoy the escapism aspect of these stories.
  • UFOs/UAPs. As the government promises to release data on UFOs or Unidentified Aerial Phenomena, scifi authors will capitalize on the public’s interest in these sightings.
  • Pandemic. After predicting this last year, I realize I was a year ahead of myself (it happens, in this prognostication biz). 2022 will be the year for pandemic scifi. Expect bizarre diseases with weird symptoms.
  • Post-pandemic. As we emerge from the COVID-19 Pandemic, scifi writers will give us tales of humanity returning to normal after devastating pandemic diseases.
  • Private space exploration and tourism. Another example of me misreading the tarot cards. It wasn’t 2021, but will be 2022 when we read scifi novels featuring billionaire-funded space travel, both for tourism and exploration.
  • Humor. We’ll see a welcome surge in funny scifi, just in time to meet the public’s need for a lighter mood.
  • CliFi.  Many readers and scifi writers share concerns about climate change, which will inspire new novels about how humanity copes.
  • AI. Artificial Intelligence will continue to prompt the scifi of 2022 as it has for years now. I predict stories involving the whole spectrum of AI, from specialized AI capable of one type of task, to general AI similar to human intelligence, all the way to superintelligence.

Personal Predictions

As a bonus, here are two prophesies involving me:

  • The North American Jules Verne Society will publish its first anthology of short stories, titled Extraordinary Visions: Stories Inspired by Jules Verne. I’m on the editorial team.
  • My collection of short stories about the future history of seasteading, titled The Seastead Chronicles, will be published in 2022.

A year from now, you’ll be amazed at how such accurate predictions were even possible. You’ll be begging for a copy of my secret scifi astrology chart and the instructions for using it. You’ll be kicking yourself for having ever doubted—

Poseidon’s Scribe

December 26, 2021Permalink

Happy Bicentennial, Frankenstein

Two hundred years ago, author Mary Shelley wrote a remarkable novel— Frankenstein, or the Modern Prometheus—which endures in popularity and bears an increasingly meaningful warning for us today.

Title page from the original 1818 edition

(Yes, I know I’m a few months late. Lackington, Hughes, Harding, Mavor, & Jones published the novel on January 1, 1818. Amazing that a publisher was working on New Years Day!)

Today, we know Shelley’s novel mainly from its numerous movie incarnations and from the term “Frankenstein monster” itself, which has become shorthand for creating something with unintended negative consequences. I’ll be commenting on the original story, though, not its later derivative works.

Boris Karloff depiction of the monster, from the 1931 movie

In my own stories, I explore the relationships between people and new technology. That is a key aspect of Frankenstein. In fact, that novel is one of the first ever to consider that theme.

Inventors typically create new technology to improve human life, to meet a need. However, the introduction of new technology can also bring about undesirable changes, including fear, active opposition, unforeseen faults in the tech (bugs), and inventor’s regret.

Not only does Shelley show us all of these aspects in Frankenstein, she turns the table on the whole technology impact concept; her sentient technology reacts to its own existence in a world of people.

To us, her novel seems well ahead of its time. Two hundred years ago, the Industrial Revolution had just begun. Electricity was a new and exciting phenomenon, not yet harnessed for effective use. Scientists were discovering elements and chemicals at a rapid pace.

Up to that time, fiction authors had written of golems and homunculi, humanoids created from magic. No stories yet existed of creating human-like life through science.

Perhaps, to readers of Frankenstein in 1818, then witnessing an explosion of scientific discovery, it might have seemed as if the animation of dead human tissue might well be next week’s news. Two centuries later, we have a better idea of how difficult the feat is. We can manipulate DNA to some extent. We’ve achieved remarkable results in extending human lifespans. We can revive the recently dead through mouth-to-mouth resuscitation and other techniques.

Mary W. Shelley

Still, we can’t do what Dr. Frankenstein did…yet. Nonetheless, when I said Shelley’s novel contains a particularly relevant warning for us today, I was referring to science’s quest to create artificially intelligent, sentient, self-aware “life.” This achievement may be decades, or only years, away. The ability for humans to create thinking, human-like life by means other than reproduction will be a breakthrough of far greater impact than any previous scientific development in human history.

We now find ourselves in the role of Dr. Frankenstein before he created the monster. We can consider the ethics of our actions in advance. We can ask if we’re insane even to pursue the enterprise. We can examine and plan for as many possible consequences as we can imagine.

Mary Shelley gave us a novel full of these consequences to consider. From twenty decades in the past, her visage warns us to be careful. She’s cautioning us with a worst-case scenario. If we fail to prepare for these consequences, we’ll have only ourselves to blame.

Thank you, Mary, for your wise counsel. On Frankenstein’s bicentennial, we’re still recklessly curious beings who discover how to do things before thinking whether we should, and before taking appropriate precautions. Maybe things will turn out fine, and much credit will go to you, for your prescient advance notice. Conveying my belated gratitude back through two centuries to you, I’m—

Poseidon’s Scribe

August 19, 2018Permalink