Everything a writer needs to know about AI (for now, at least)

Reading Time: 6 minutes

I’m writing this blog post from Lima, Peru, where I gave a keynote speech to a packed room about generative AI, its risks, benefits, and what’s going to happen next.

Maria Korolov giving the keynote presentation at the Data & AI Summit. (Image courtesy Seminarium Peru.)

Oh, you can’t really see me up on the stage in this photo. Here’s a closer one:

That’s me up there. (Image courtesy Seminarium Peru.)

I was invited here because I’ve been writing about AI at my day job for several years now — and about generative AI in particular these past couple of years. I’ve spoken to hundreds of experts — business executives about how they’re using generative AI and where they’re seeing benefits and where it doesn’t work — as well as to analysts, consultants, and vendors. I have a broad and neutral view of what’s going on.

I’m not trying to sell anything to anybody. I just want to know what’s really going on with AI.

Plus, as a writer and journalist, I have a personal stake in figuring out what’s happening because I have a strong feeling that my whole career is on the line.

The story so far

In 2022, we saw the mainstream emergence of image generators like DALL-E 2, Midjourney, and Stable Diffusion. ChatGPT came out at the end of 2022, just in time for the holidays.

In 2023, we had more text generation AIs, plus code generation and audio generation. We saw deep fakes that were indistinguishable from real humans. This year, we’re seeing video generation that’s almost as good as human video, 3D graphics that will transform video games, and autonomous agents that can create plans, delegate tasks, and do other higher-order jobs that simple ChatGPT can’t do.

The speed at which this technology is evolving is unprecedented. I’ve covered many transformative technologies in my career, but generative AI is moving faster than anything I’ve ever seen. To give you a taste of how fast it’s moving, in March 2022, Midjourney’s attempt at drawing a kitten was laughable.

(Images by Maria Korolov via Midjourney.)

Just eight months later, the same “kitten” prompt produced stunning results.

Generative AI is now winning art and photography awards, with even the best experts unable to tell it was AI-generated. This year, generative AI is helping win literary prizes.

Companies are investing heavily in generative AI, with 93% of CEOs either already investing or planning to do so. The return on investment is insane — according to an IDC survey of over 2,000 companies, for every dollar invested in generative AI, companies are seeing an average return of $3.50. Experts say that while computer power typically doubles every two years, per Moore’s Law, with generative AI, it’s doubling every three to six months.

Why should writers and artists care about what companies do? Because companies are spending very, very large sums of money on AI. That means even more investment, more technological development — and lots and lots of pressure on regulators to allow it to keep happening.

The next big breakthrough is going to be world simulation, where an AI can simulate the entire world — text, pictures, audio, video, together with all our scientific knowledge and raw research. This is going to be a major step towards general artificial intelligence, which is AI that has common sense and is as smart and flexible as a human.

What about creativity?

But is generative AI just autocomplete on steroids, cutting and pasting stuff from the internet without real creativity or original insight?

I wish.

Yes, early machine learning was just predictive analytics. Using statistics to figure out what comes best. But today’s generative AIs look for patterns behind the patterns, and patterns behind the patterns behind the patterns. That’s getting us pretty close to what looks like understanding.

And if it looks like understanding, walks like understanding, quacks like understanding… well, for people who are buying this stuff, that can be plenty good enough.

I did my own test with Claude AI late last year, uploading an unpublished novel and asking it to identify the themes, character arcs, and plot holes.

Sometimes people think that when you ask an AI to analyze a book, it just spits out a summary of other people have said about the book.

But nobody has ever said anything about my book because it’s not published. Nobody’s read it yet, except for a few beta readers and an editor, and they didn’t post anything about it online.

Claude AI read my book in seconds, accurately identified the themes and character arcs, and found plot holes that even my human readers and editor had missed. It then offered suggestions on how to fix them in a way that fit with the book’s themes and character arcs. This is not just cutting and pasting — it’s understanding and analyzing. Or close enough, anyway,

Oh, yeah, and in a study where humans and generative AIs were given creativity tests, the AIs did better than the humans.

Other than divine inspiration, humans do three things for creativity — they extrapolate from things, they combine different ideas into something new, and they use randomness — Tarot cards, writing prompt generators, random pages in encyclopedias, writing down their dreams, or doing drugs.

AIs are good at extrapolating, good at combining, and are great at coming up with random stuff.

Want to see it in action? Use this prompt: “Please give me three random ideas for sci-fi novels, combining disparate concepts not usually found in sci-fi! Thank you!” Yes, it gives you better results when you’re polite. It’s one of those quirky AI things.

Are you back? Are you terrified yet? I certainly was. I could see my entire future disappearing. I’m too young to retire!

But after the panic faded, I realized that things aren’t nearly as bad as I thought. Humanity had lived through other technological revolutions before, including ones where we got dramatically more productive. Remember when we all used to be subsistence farmers, constantly living under the threat of famine or plague? Today, only 11% of the world’s population is severely food insecure, and that’s because there are still countries that haven’t fully gone through industrialization yet. Plus, when things get bad, the rest of the world pitches in with food aid and medical relief, provided local governments don’t keep the supplies out.

Only a few of us need to work in agriculture in order to feed everyone else. But the rest found other stuff to do. We became teachers, doctors, professional athletes, astronauts, programmers, scientists, TikTok influencers and Twitch streamers — we wouldn’t even be able to explain most of our jobs to people from 200 years ago.

AI will increase productivity, just like tractors did. With more productivity, we’ll be able to do more stuff. A lot of that “more stuff” will be in the form of new types of work that we can’t even imagine yet.

But say you want to stick with the work you’re currently doing.

I mean, I like being a writer. I don’t want to go and be… a personal motivation consultant who beams her image directly into people’s brains? Or whatever it is we’re going to do next.

What can we do?

So, how can writers and other creatives survive when AI can do everything we do?

We can just look at other industries that have been disrupted by technology and survived, like small farmers who leaned into being organic, free-range, and locally sourced. The key is to promote these qualities — if you don’t tell people your apples are free-range and organic and anti-biotic free, they’ll just think that they’re uglier, more expensive versions of mass-produced ones.

The bottom line: don’t compete directly against the technology. It’s going to be faster and cheaper, and there’s no point in getting into that race. Instead, compete on the fact that you’re human.

Look at Taylor Swift — even though music is virtually free these days, she still makes money because people care about her personally. They buy special editions, all her merchandise, and concert tickets because they want that human connection. They feel that she’s speaking for them.

If you’re a writer, artist, or editor, lean into that human connection. Interact with your readers, develop personal relationships with them, and offer them opportunities to support you via Patreon, merchandising, and in-person or video appearances.

If you don’t know how to start, join us on our Free Friday videos, where we talk about fun free sci-fi and fantasy bestsellers. Readers can get to know you by finding out what you like and don’t like. If they like your taste, they might check out your books.

Just email me at [email protected]. We’re also looking for columnists, book reviewers, and editorial volunteers.

And if you want to get a sense of me as a person, watch the video below:

MetaStellar editor and publisher Maria Korolov is a science fiction novelist, writing stories set in a future virtual world. And, during the day, she is an award-winning freelance technology journalist who covers artificial intelligence, cybersecurity and enterprise virtual reality. See her Amazon author page here and follow her on Twitter, Facebook, or LinkedIn, and check out her latest videos on the Maria Korolov YouTube channel. Email her at [email protected]. She is also the editor and publisher of Hypergrid Business, one of the top global sites covering virtual reality.

4 thoughts on “Everything a writer needs to know about AI (for now, at least)”

  1. I love this: “I don’t want to go and be… a personal motivation consultant who beams her image directly into people’s brains”…

    and while I am a writer, too, and I don’t want to use AI or have it use me, you are so right, Maria–we can’t compete in a direct assault. Plus, as you say, no matter what has developed, thus closing some doors, it opens new doors we can’t even begin to imagine right now. Every single time the nay-sayers have said the world will come to an end. And it doesn’t. Creativity is of the spirit, not only the mind.

    I am wary of an AI take over, though! Too many movies I’ve seen on the subject? Your comment about its increasing sophistication was not a consoling thought! But I am also a Star Trek fan and they worked over the effects of AI in a dozen ways, all of which turned out well either because life really was enhanced or the human element became uppermost even in the midst of high tech. They did, after all, find a way to surpass the Borg.

    Anyway, thanks for the great blog and also video. Plus kudos on the Peru presentation. Wow.

    Regina Clarke

      1. Oh, and I have posted a video based on my Peru presentation on my own channel:
        https://www.youtube.com/watch?v=NfFBzjFMDaI

        It’s not *exactly* the same — in Peru, I referred back to previous presenters, sat down for an on-state Q&A with the conference host, spent more time on some of the more technical and business-y topics, but the slides are the same, and I covered pretty much all the same material.

Leave a Comment

Your email address will not be published. Required fields are marked *