A deep dive

'As much mud in the streets as if the waters had but newly retired from the face of the earth, and it would not be wonderful to meet a Megalosaurus, forty feet long or so, waddling like an elephantine lizard up Holborn Hill.'

That's an excerpt from Bleak House, and has nothing to do with this week's post except that it perfectly evokes a diluvial theme and it's an example of masterful writing. We'll get back to that.

A partially flooded world is the setting of Bellwater, a tabletop roleplaying game about salvage crews operating in an inland sea created by a cataclysmic deluge centuries earlier. The game focuses on themes of debt, breath, and the recovery of relics from drowned cathedrals. What makes Bellwater unique is that it was written by OpenAI Codex. Ethan Mollick, who created it as an experiment in current AI capabilities, says: 'The setting is interesting and novel, and the rules appear to make sense, drawing on existing game patterns while adding unique elements.' 

I agree. I really like the game concept and the vaguely Edwardian feel of the setting. Inspired by Simon Stålenhag's Things from the Flood, you think? Possibly, but quite obliquely and among many other sources that might well include (AI being well-read if nothing else) Charles Dickens, J.G. Ballard and Marcus Sedgwick. The Old Testament swiped from the Epic of Gilgamesh, after all, so under the sun there is no new thing.

Professor Mollick does go on to point out the flaws: 'If you are a frequent reader of AI writing you see the same problems here: a love of the uncanny; overly complex ideas that do not fully pay off; weird metaphors (“weather and architecture are the same argument at different speeds”); too many ornate sentences (“the holy things that surface when a sea forgets it was once a road,” is cool once, an entire book of that is exhausting); dialogue where every character speaks in the same clipped tone.'

There's a lot of stinkingly bad writing out there, both AI-created and by human authors. Some of the latter is very popular, so it can't be that readers disdain slop. The difference is that the AI is getting better month by month, whereas the humans who are writing execrably aren't even trying to improve their craft (or their grammar). No author should just tell AI "write a book", but it can be very useful in brainstorming concepts, researching obscure details, and maintaining a world bible. Podcaster Joanna Penn explains here all the ways that authors can use AI as a useful tool of the craft, and director Ash Koosha talks about how he's used AI in filmmaking here.

If you want to take a look at Bellwater, Prof Mollick has made it available to download and has been revealing some of his experiments with early access to Claude Mythos. I'd be interested to hear from anybody who actually tries playing a game of Bellwater. You may not feel that it is done well, but reflect on what Dr Johnson said and be surprised to find it done at all.

Seminar: "The AI Assisted Artisan Author"

The main reason I abhor terms like "AI slop" is that they are propaganda, not arguments. They are used as substitutes for thinking and reasoned debate. (It's the same mindset that coins slurs like "Crooked Hillary" and "Comrade Kamala"; "stochastic parrot" is another.) The fact is that some things generated by AI are slop but many are not - in physics, biochemistry, medicine, etc. Only a fool, given that the genie is out of the bottle, refuses to engage with it. Also there is, and always has been, plenty of slop that humans created for themselves without the help of AI. Just cast an eye over the bestseller lists.

AI art comes in for a lot of flak, not entirely unfairly. It's much more polished than anything I could draw. But it's only workmanlike; it's never great. The people who are quick to coin terms like "AI slop" have latched onto the claim that using AI art is putting human artists out of work. I can't speak for others, but if a project of mine has a budget then I want to work with human artists every time: Inigo Hartas on Jewelspider, Leo Hartas on Mirabilis: Year of Wonders, Mattia Simone on Vulcanverse, Russ Nicholson and Kevin Jenkins on Fabled Lands. I make sure they get paid even if (as often happens) I don't. No AI can do what they can do.

A lot of projects don't have budgets, though. Blog posts, for example. For a decade I'd have to trawl through mediocre public-domain images to find something to illustrate the week's piece. Now, in the absence of anything better, at least Gemini can whip up something passable. After spending an hour or two planning and writing a post, not having to waste half an hour scouring the internet for images is a godsend.

Most gamebook and RPG authors don't make any money. Their works are labours of love. After months of writing, when they were finally ready to publish, it used to be that their only option for illustrations was to find some out-of-copyright art. That could occasionally be just right -- who else but Gustave Doré could illustrate James Wallis's storytelling game The Extraordinary Adventures of Baron Munchausen? More often than not it was just the creator making the best of a bad job. They weren't putting a human artist out of a job, any more than you're paupering a plumber or carpenter when you do some DIY based on a YouTube instructional video, because there never was a job there in the first place. 

And incidentally there'd be no indie game scene at all if not for desktop publishing software and print on demand. Those came along more than twenty years ago. Do you see the streets lined with out-of-work typesetters and printers? No, and this technological revolution won't wreck anyone's career either.

So much for art. What about writing? AI is really not good at fiction. Its prose (learned from humans, of course) is ungainly. That would be hard to fix because elegant prose is not easily evaluated and so it would be hard to train a model to know the difference. It's not like coding or maths, where there's a clear difference between a right and wrong answer. (That's why, when teachers had to estimate their pupils' exam performance during the covid lockdown, there was much greater variance between the estimate and the eventual exam result for arts subjects than for sciences.)

Even if we could teach an AI to write beautifully, it couldn't (currently, anyway) write a good novel because it has no depth of insight -- though, again, that's true of many human authors. The telephone-directory-sized romantasy bestsellers stacked up in the bookstore window are not good novels, just popular ones. Lee Child's working method for writing the Jack Reacher books is effectively just what an LLM does, so there's no reason ChatGPT couldn't come up with a passable pastiche of one of those. But it's not going to rival Flaubert or Turgenev for a while yet. Owning a camera doesn't make you Da Vinci, after all.

I realize that in the 2020s there is no longer any possibility of convincing anybody that they might be wrong or even that somebody who disagrees with them isn't a knave, but maybe we can all still agree that it's better to be informed than not. That's why I'm recommending Joanna Penn's online seminars on ways to use AI to make you a better author. For the reasons given above, that doesn't mean prompting it with, "Write this novel for me." What Jo is covering includes the background tasks: deep research, brainstorming and ideas, outlining, structuring, plotting, and planning, characters and worldbuilding. For example, she explains how to use NotebookLM to maintain a world bible. (That would have saved months of wading through texts if it had been around when I was working on the Lyonesse RPG, for example.)

There are seminars on 16 and 23 May. I've taken Jo's seminars before -- even after 40+ years as a working author there's plenty I can still learn -- and they are worth every penny. Get your tickets here.

An audience of one

There's Matthew Berman reminding us that future is coming up faster than you think. He's talking about videogaming, but the same principles apply to movies, comics, and literature.

The novel – at least, the genre novel – may well go the way of the epic poem, to be replaced by something more like an RPG session which an AI will run for the reader. (Or, more likely, the listener or viewer.) The top authors will devise the elements of the story, the characters and timeline (perhaps more like creative directors than old-style authors) and the AI will use that to tell a story that gives prominence to the bits that interest the individual reader. Did your parents make up stories to tell you when you were little? Like that. Or maybe like this.

You'll still discuss the story with friends (an important feature of most entertainment) but the specific events in your version may vary from theirs. Initially such on-the-fly stories will be trite because roleplaying has been infected by a lot of Hollywood pablum about act structure and story tropes, and that’s what the AI models will learn from. But eventually it may shake that off and become a new independent art form. "Not a line, but a bolt of lightning," as C W Longbottom puts it:

In the meantime, a market will remain – small, though, and shrinking – for grown-up fiction that doesn’t pander to YA tastes. Genre fiction falls in predictable patterns involving plot, and so is easily copied by novice writers and neural nets, whereas literary fiction is harder to fit to a formula because it usually concerns itself with the unique outlook and choices of the characters. But don't assume that because the AI hasn't experienced human emotions it won't eventually be able to write Lolita or War & Peace. Conrad didn't personally have to hack his way through an African jungle to learn how to write Heart of Darkness. It's only a matter of time before those more complex story patterns are learned and replicated by AI, just the same way that most authors do it. And then we'll be in a whole new world of entertainment.

Interview on The Creative Penn podcast

It was an honour and a pleasure to be invited recently onto Joanna Penn's podcast The Creative Penn. We talked about games, comics, books, publishing, story worlds, and how writers can use AI tools in their work. One word of warning: Jo has 810 episodes in the back catalogue, and with her justifiable reputation as a visionary who is often years ahead of her time, you might feel compelled to listen to the whole lot. And as well as being a podcaster, entrepreneur and commentator on the publishing industry, Jo is an award-winning author with several successful series to her name, so be sure to check out her books too.

The next wave is the big one

2010. Leo Hartas and I had tickets to the London Book Fair. We weren't expecting much to come of it. If an author steps onto a publisher's stand at the LBF, they look at you like they want to call RentoKil. But we had just published our comic Mirabilis: Year of Wonders on iPad and we were hopeful of showing it to other writers and artists at least.

Fate smiled on us, though. A volcano erupted in Iceland, disrupting air travel, which thinned out the crowds at the LBF. Deprived of the international reps to do deals with, UK publishers had nothing better to do with their time than hold their noses and talk to the authors. (Oh, you thought they'd welcome the opportunity, seeing as we're the people whose work pays their salaries? Ha ha.)

Still, Leo and I thought we had something pretty interesting to show them. We knew they'd have zero interest in the creative content of our Mirabilis app, but there were features they should be taking note of. "Look at how you can buy each issue of the comic in the app," we said. "Those could be the latest titles in your book list. There are share buttons, and if the user signs up they'll get told about new titles in any series they're following."

The publisher's eyes scanned the crowds behind us, probably wondering where that exterminator had got to. "We are publishers, not booksellers," she told us with infinite disdain.

"But this would give you a direct relationship with your customers. You can find out what they like. Sell directly to them. Push additional content like author interviews. Let them know about upcoming releases. All within the app."

She turned away, visibly sickened by having to talk to tradesmen. Mama would never have had to stoop so low. "It is not our business to have a direct relationship with our readers. Publishers do not need 'apps'."

Fast forward only six or seven years and every publisher by then employed bright-eyed tech advisers. "What we want as publishers," they would tell you loftily, "is a direct relationship with our customers."

Too bad they didn't get on that bandwagon when it was setting off. Perhaps if Leo and I had worn suits and pretended scorn for the saps who wrote and drew the comic, the publishers at the 2010 LBF might have listened. But probably not. It's an unchanging trait of British publishers that they will completely refuse to embrace any new trend till it's already passed them by. (A few years earlier, discussing ebooks with another publisher, they'd asked us where they would sell "the discs" with the ebooks on.)

So I'm wondering what trend they're missing right now. It might well have to do with AI. Here's a podcast about that very point featuring Joanna Penn and Thad McIlroy. And if you find that interesting try the episodes on writing with generative AI and using generative AI in book cover design.