My sweet Claude

I'm a technophile through and through, but even I know that not everything that can be done should be done. Case in point: there's a service where you can order up a plot and characters and have a bespoke novel written and printed just for you.

(Hang on a tick, though. We'll be talking about AI here, and I can't be doing with the obligatory two minutes' hate of people telling me their tribal identity via a series of muddled anti-AI arguments, so take a look at the blog policy on that one, and if it's not to your taste then come back tomorrow.)

OK, so this customized novel service involves 'romantasy' books and the way it works is each reader specifies their required hero and heroine type, the degree of steaminess, whether the relationship should be bittersweet or rosy, and any necessary plot details. Then an AI writes the thing, the company prints a copy, and it's yours for a mere $89.

I have no doubt the results are as horrible as they are expensive. There is quite enough human-generated slop in publishing without pressing a button and having machines add infinitely more. The AI models get grammar and spelling right, and their prose style is marginally better than some bestselling authors, but they still shouldn't be allowed anywhere near the actual authorial process. For one thing, speaking as a writer, the words are the fun part. Why would I want anyone or anything else to do the writing for me?

I do use AI every day, though, in the tedious support tasks for which, if gamebooks and roleplaying games made any money, I'd hire an assistant. This month I'm making a push to complete my pet project, Jewelspider, a book which is now over two hundred pages long and three years late. I'm drawing on a host of primary sources, many of which are in Latin, French or Middle English. Finding all references to medieval lycanthropy, say, in a few dozen books like that could eat up weeks -- or I can get Gemini Notebook to help me and have those weeks free for writing instead.

In the middle of that, I'm still fielding the dozens of micro-tasks that would suck up time if I let them. A publisher needed me to give them a list of fifty scenes to illustrate in a couple of gamebooks I wrote  years ago. I'd have to re-read the books, find fifty key scenes, and collate all the text the artist will need across several sections. That's a few days of the most mind-numbing work... for a human. Claude read the books, found fifty different dramatic scenes, made sure to avoid spoiler scenes, and from my text pulled together a briefing for the artist for each illustration. I was able to check that list in an hour, make a few corrections, and it was good to go.

Then an app developer who is considering adapting another series of mine asked if I could provide them with a list of every item that can be picked up in the books, where they can be used, what they do, and where the player can lose them. That's a good month of work as the books were written decades ago and I don't still have the flowcharts. But Claude did the job in twenty minutes and gave me all the info in an Excel spreadsheet. It was two hundred items, incidentally. If I'd done that myself I'd still be working on it now.

Another app developer is producing an audio version of my gamebook Heart of Ice. Every few days they send me the scripts. That's dozens of pages to check, not only for dialogue but for the logic markup that drives the game. If I allowed myself to get pulled down that rabbit hole, Jewelspider would never get finished. So Claude took a look and came back with a list of the things I needed to take a closer look at -- dialogue given to the wrong character, inconsistent tagging, that kind of thing. Another couple of weeks saved. When the developer emailed to say they'd be recording the next day and the actors needed a pronunciation guide to all the items, names, places, etc, in the book, it could have meant me pulling an all-nighter. Claude took ten minutes to find all the words that might get mispronounced along with suggested alternative ways to say them, so all I had to do was tell Claude which were right and I got a list back with both a plain phonetic respelling and (for especially diligent thespians) an IPA version of each word.

The AI dealt efficiently with chores that once would have cost me time and energy, of which I nowadays have far less than I did in my youth, allowing me to focus on the creative part of my job: the actual writing. And if you're interested in Jewelspider (which is all-human work, both the writing and the art) please come over to Patreon and take a look.

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 ulcer on the big toe

‘People whose minds are—like [Arthur] Machen’s—steeped in the orthodox myths of religion, naturally find a poignant fascination in the conception of things which religion brands with outlawry and horror. Such people take the artificial and obsolete concept of “sin” seriously, and find it full of dark allurement. On the other hand, people like myself, with a realistic and scientific point of view, see no charm or mystery whatever in things banned by religious mythology. We recognise the primitiveness and meaninglessness of the religious attitude, and in consequence find no element of attractive defiance or significant escape in those things which happen to contravene it. The whole idea of “sin”, with its overtones of unholy fascination, is in 1932 simply a curiosity of intellectual history. The filth and perversion which to Machen’s obsoletely orthodox mind meant profound defiances of the universe’s foundations, mean to us only a rather prosaic and unfortunate species of organic maladjustment—no more frightful, and no more interesting, than a headache, a fit of colic, or an ulcer on the big toe. Now that the veil of mystery and the hokum of spiritual significance have been stripped away from such things, they are no longer adequate motivations for [fantasy or horror fiction]. We are obliged to hunt up other symbols of imaginative escape—hence the vogue of interplanetary, dimensional, and other themes whose element of remoteness and mystery has not yet been destroyed by advancing knowledge.’

We're almost a hundred years on from when H P Lovecraft wrote that and still plenty of humans imagine the universe being ruled over by a stern parent with a set of rules and punishments for naughty children to fret about -- and no two groups can quite agree on what the stern parent's rules are. It's pretty much the textbook study in how children grow up emotionally troubled. At this point I just hope we don't infect the AGIs of the future with our ape-brained notions.

HPL was a tireless champion for the sense of wonder. He was simply opposed to the category error that puts the numinous in the same box as objective reality. This is from the introductory paragraphs of  'Supernatural Horror in Literature':

'Relatively few are free enough from the spell of the daily routine to respond to rappings from outside, and tales of ordinary feelings and events, or of common sentimental distortions of such feelings and events, will always take first place in the taste of the majority; rightly, perhaps, since of course these ordinary matters make up the greater part of human experience. But the sensitive are always with us, and sometimes a curious streak of fancy invades an obscure corner of the very hardest head; so that no amount of rationalisation, reform, or Freudian analysis can quite annul the thrill of the chimney-corner whisper or the lonely wood. There is here involved a psychological pattern or tradition as real and as deeply grounded in mental experience as any other pattern or tradition of mankind; coeval with the religious feeling and closely related to many aspects of it.'

If that's the kind of fiction that sets up a stirring in your soul, you might like some of the offerings in Wrong magazine: stories that take the reader over the boundary into a place where everyday reality meets the inexplicable. Cue the music.

*  *  *

(While we're on the subject of Lovecraftian horrors -- if you missed the funding campaign on Gamefound for Whispers Beyond The Stars, there's now the opportunity to secure a copy with a late pledge. But don't dilly-dally. I can't guarantee there'll be a third chance.)

Canaries in a lunar mine

At one point in Cthulhu 2050: Whispers Beyond The Stars, the protagonist (that's you) might need to travel to the Moon. There are extensive lunar colonies with about ten thousand people living there, but they’re almost all working in highly specialized technical jobs and a ticket to the Moon isn’t cheap. So how are you going to get up there?

My process of collaboration with Paweł Dziemski was that I wrote the first draft, comprising about half the sections, and Paweł then expanded that and in doing so he had to fix all the loose ends and plot holes that I’d blithely left in my wake. (See how sneaky I am?) So I pondered what skills the protagonist could offer to convince a corporation to pay for them to go to the Moon – assuming that by that point in the game you haven’t saved enough credits to buy your own ticket and aren’t stealthy enough to stow away. One possibility was this:

I’d been focusing on how to make colonies in space credible. In short, if you have super-intelligent robots, why send humans at all, other than maybe a few supervisors? One job I figured that robots couldn’t do to everyone’s satisfaction was sex work*. Most humans will probably continue to prefer screwing a partner with no nuts and bolts. Well, with no bolts, anyway.

I was thinking of the SF I read back in the early 1970s, when the Western world was a lot less prudish than today. But Paweł gave me a reality check, pointing out that a digital gamebook might be played by children, and (again, unlike the ‘70s) parents get very prissy about controlling what their kids see and read.

I hadn’t thought of that. On an episode of his Cautionary Tales podcast, Tim Harford talks about Alvin Roth’s theory of repugnant markets. Repugnant markets include trade in kidneys and (in many countries) prostitution. Roth solved the kidney transplant problem with the concept of an exchange – if two people in need of the service (a transplant) both had a willing donor (say a relative) whose tissue type was incompatible with them but would work for the stranger, the market could put them in touch with each other. The equivalent trade in a society that regards prostitution as a repugnant market would be partner-swapping parties or orgies – effectively an exchange in sexual services.

“A profession that is always violent at its core,” is one common view of prostitution. That’s similar to the argument that used to be heard in the 1960s that homosexual politicians were more likely to fall prey to foreign spies. Such politicians were more at risk of blackmail, but not because of their innate character and sexual preferences; it was because society at the time marginalized and even outlawed those preferences. It was a self-fulfilling prophecy.

I’d anticipated a future in which prostitution was legal, regulated, and therefore no longer controlled by human traffickers and violent criminals. But that’s a 20th century liberal view of the future, not the thinking of a culture now dominated by American puritanism. Paweł was quite right. We didn’t want irate parents and fundamentalist religious groups creating a fuss, so we quietly shelved the sex-work option and went for this instead:

That’s way better. When you have to abandon your first idea, the second thought is usually a big improvement. Murder your darlings, as Sir Arthur Quiller-Couch said. My version was merely satirical: the best job an underskilled human can hope for in a fully automated future is hiring out their body for sex. Paweł’s revised version is genuinely chilling: humans are used for dangerous work because they are cheaper to replace than the top-quality robot models. We got there by a process of collaboration, which proves that two heads are better than one – and not just in bed.

* Nick Bostrom suggests two other professions he thinks people won't want robots/AIs to do: priest and politician. He could be right. I find it hard to believe there are still priests in the 21st century, but then it's not the 21st century I signed up for. As for politicians -- if AI couldn't do that job better, it really isn't worth bothering with.

Forget the thumbscrews

A trope you often see in TV shows and movies these days is the hero torturing a suspect to get information. It cropped up in Watchmen and Daredevil among many others. Sgt Sam Provance, who exposed abuses at the Abu Ghraib prison in Iraq, points out in this BBC interview (25m25s in) that many of the real-life torturers in the US Army grew up thinking that's what the good guys do. An audience’s desire for a “strongman” (or in Watchmen’s case a "strong" woman) to brutalize the bad guys usually comes at times when people feel powerless and frightened, a sure sign of a civilization at a low ebb.

A truly strong hero in a confident democratic society doesn’t need to behave like a sadistic bully. I’m thinking of the Allied officer after the liberation of a POW camp in Burma. The Imperial Japanese commanding officer came to see him and asked why the Allied troops weren’t starving and beating the Japanese prisoners, given the way the Japanese had maltreated their own POWs. The Allied officer replied, “Because we’re better than you are.”

Torture is also usually ineffective at getting reliable information, though admittedly the states that employ it don’t usually care much about the truth. But we’re not here to discuss the effectiveness of torture, nor the libertarian psychology that stokes fantasies of it. (Hollywood writers have been embedding libertarian ideology into their scripts for decades, after all. I'm sure if you asked the President of the United States he'd yap, 'Torture? Big fan, big fan.') I'm not even going to talk about the moral arguments, which you'd hope wouldn't have to be explained to any civilized person. My gripe about torture in story terms is that it’s just plain boring.

Real interrogation, whether in fiction or roleplaying, gives you an opportunity for an interesting scene. One character has information they don’t want to reveal. The other character needs to gull them into telling the truth. Like so:

In GURPS, Interrogation is a skill quite separate from Intimidation, though the GURPS designers make the classic mistake of giving a huge Interrogation bonus for the use of torture. Players therefore tend to get the pliers out and make like Beria’s thugs, which is a pity because role-playing the scene in which the interrogator tries to outwit their adversary is potentially way more interesting than any combat could ever be. (To get information in the field, use elicitation.)

The examples of torture in Daredevil are particularly uninspired because DD has special abilities that help him to tell when somebody is lying. But come to that the writers also gave him a gruelling concussive battle with a dozen or more mobsters where he could just have pulled the fusebox off the wall and beaten them easily in the dark. So it’s not too surprising that the writing of other scenes was sloppy and ill thought-out.

There are writers who do it well. Try the novels of John le Carré, whose interrogators employ the gamut of psychological tricks -- cajoling, bullying, charming, flattering, coaxing, probing and misleading until their subjects reveal the information, often without realizing how much they've given away. Le Carré's interrogators never need to dangle somebody off a roof or pull out their fingernails, and those scenes are so much more gripping because of that. (For example Tinker, Tailor, Soldier, Spy or A Perfect Spy.) Or consider Geoffrey Household's classic thriller Rogue Male, where the antagonist extracts a confession from the narrator who has not previously admitted the truth even to himself -- and that's under duress but without the use of physical torture, which is used by the bad guys right at the start of the novel with no useful results whatsoever.

Then take a look at this video in which a real-life expert interrogator analyses movie writers’ ideas of how it works. And next time you have an interrogation scene in a roleplaying game, consider how much more interesting it is to have it play out sans the waterboard and the big-dick posturing.

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 same cigarettes as me

So I'm not the only one irked by seeing modern characters popping up in period dramas. I don't know who writes the Movie Media Hub posts on Facebook, but I'm with them on this point. Not specifically feminist characters, though: any character whose function is to spout modern attitudes a few centuries ahead of their time.

I disagree about why such characters are inserted into stories. There are multiple reasons. Most obviously, it's a way for a lazy writer to get laughs. "Heh heh, it's Victorian times but Holmes just said 'OK Boomer'." Ursula K Le Guin griped about fantasy writers who put contemporary American slang into their heroes' mouths:

"...Since fantasy is seldom taken seriously at this particular era in this country, they are afraid to take it seriously. They don’t want to be caught believing in their own creations, getting all worked up about imaginary things; and so their humor becomes self-mocking, self-destructive. Their gods and heroes keep turning aside to look out of the book at you and whisper, “See, we're really just plain folks." '

It happens because some audiences are easily pleased. They are asking, as Oscar Wilde put it, for art "to please their want of taste, to flatter their absurd vanity, to tell them what they have been told before, to show them what they ought to be tired of seeing, to amuse them when they feel heavy after eating too much, and to distract their thoughts when they are wearied of their own stupidity."

Anachronism is also a quick and easy way to get the audience on-side with a character. How can you not root for the guy who's denouncing slavery in the 18th century? Well, there were indeed plenty of people denouncing slavery back then, and a good thing too, but if you are going to put them in a script they shouldn't talk like 21st century characters. Unlike us today, they exist in a world where it apparently wasn't obvious to everyone that enslaving another human being was a monstrous crime. And even if you get the mindset and the dialogue right, it's still a lazy way to create rooting interest. Consider first whether you need to start with a hero whose every attitude is right-on. We were talking about the superfluity of likeable characters in a recent post and, as a great storyteller once pointed out, it is more satisfying to see one sinner that repenteth than ninety-nine decent characters who need no repentance.

Mainly, though, I detest this trope because of its cultural chauvinism. If somebody is interested in a story with an historical and/or non-Western setting, what is the point of inserting modern Western characters? You may say we can't fully know how people in such settings would see the world. Equally, we can't all be as selfless as the Buddha but that doesn't stop us giving money to charity. We make an effort at all sorts of things we can't hope to do perfectly, so why not storytelling or roleplaying? We do know that a farmer in medieval Iceland won't share the attitudes of people at a TED talk. If we want the farmer to make a stand for something we believe in today, we should look for a way he'd credibly conceive of and frame that in the context of his times. In writing or playing such a character, we can find stepping stones that help us imagine their interior life, for example using a process that Robert Graves called analeptic mimesis, effectively taking what you know of a time and place and dreaming it back into existence. That is true even in a fantasy story, as long as it's a fantasy setting where somebody has thought out some of the culture and economics that shape it.

In his book Medieval Horizons, Dr Ian Mortimer contrasts historical accuracy with historical authenticity:

"We impose our own prejudices on the past and reinvent it as 'how it must have been' to conform with our outlook. To appeal to a modern audience, therefore, a medieval heroine in a book or film must be shown as having control over her own life. Alternatively, she must appear constantly fighting her oppressors. Male peasants must similarly include at least once plucky, Robin Hood-like character who leads the others in defying authority. The result is a reproduction of modern society draped in medieval clothing." 

It might be an interesting exercise to write a modern drama but insert one character with the (imagined) attitudes of somebody from the 23rd century. Presumably even the least demanding audience member would spot something amiss then. And how would they take to having their currently fashionable beliefs scorned as antediluvian and even toxic by some supercilious git from the future?

Of course, your tastes may vary. Looking for fiction or games set in the early 19th century, do you want Middlemarch and War & Peace -- or, alternatively, are you looking for Bridgerton? Taking an imaginative leap into another society, or just dressing-up in period costume? Now you know where I stand. Still, those Regency greatcoats are jolly tempting.

Automatic writing

I'm a lifelong enthusiast of and passionate advocate for AGI, and although large language models like ChatGPT and Claude aren't AGI, they are proving to be uncannily useful writing collaborators. I should qualify that. Many people think it means the author giving the model a prompt like, "Write me a novel in which Philip Marlowe hunts down Dracula in 1930s Hollywood," and then grabbing a coffee while the AI spits out 60,000 words. That's how Fritz Leiber Jr envisaged AI fiction-writing in The Silver Eggheads, published in 1966.

Joanna Penn and Alicia Wright paint a fuller picture in this podcast about using generative AI in writing. Every writer finds their own way of making best use of the technology. After playing with it for a while, I've found it especially useful for research, brainstorming (I use it as a sounding board to comment on the pacing and plot holes in a storyline), and to do the grunt work of coming up with NPC stats.

Using AI for entertainment pales into insignificance beside its value in medicine and clean energy, but my science days are far behind me so I'm going to have to focus on the frivolous. For an idea of how LLMs can help with writing research, take this example of a conversation with Claude when I was working on a Faerie Queene scenario for Patreon:

Dave: "Can you give me any examples from late medieval or Tudor literature of a character posing a conundrum that tests the wisdom and moral insight of the listeners? Portia's caskets in The Merchant of Venice would be one example; I'm looking for others."

Claude: "There are several fascinating examples from this period. Let me share a few notable ones. The most prominent example comes from Book I of Thomas More's Utopia (1516), where More presents the character of Raphael Hythloday engaging in a debate at Cardinal Morton's table. Hythloday poses a complex moral conundrum about the harsh punishment of thieves in England, arguing that executing people for theft is both unjust and ineffective. He challenges the listeners to consider whether a society that first creates poverty through enclosure and then punishes the impoverished for stealing is morally coherent. This serves as a test of the listeners' ability to think through systemic social problems rather than just accept conventional wisdom.

"Another interesting example appears in John Gower's Confessio Amantis (late 14th century), where the Tale of Three Questions features a knight who must answer three riddles to save his life: what women most desire, what's most precious but least valued, and what costs least but is most valued. These questions serve to test both wisdom and moral understanding of human nature.

"In the morality play Mankind (c. 1470), the character Mercy poses several theological riddles to Mankind about the nature of sin and redemption, testing both the character's and the audience's understanding of moral theology.

"An unusual example comes from John Skelton's Bowge of Courte (c. 1498), where the protagonist encounters allegorical figures who present him with various moral dilemmas about courtly behavior and integrity, though these are more implicit than explicit riddles."

Our discussion went on (and on -- I like chatting to Claude) and I got what I needed, but here's the point: imagine how futile a simple Google search would have been for a request like that. Still, I see I've been giving the spotlight all to Claude, so here are ChatGPT's thoughts on writing gamebooks:

ChatGPT: "Since a 1000-section gamebook requires long-term consistency, custom formatting, and mechanical coherence, GPT-4-turbo or Claude 3 would be your strongest options. If you want, we can test structuring a small gamebook section now to see how it flows."

Dave: "OK, let's try it. Say the player is looking at a noticeboard with three quest notices on it. We'll use keywords, and each notice has two keywords associated with it. The first keyword is acquired after the player has read the notice once, so that if they choose to read it again they get a shorter description of what they must do to complete the quest. The other keyword in each case is acquired by completing the quest (which is outside the scope of this example, it's something they would go off and do in another part of the game world) so that if they look at a notice for a quest they already did, they are reminded that it's no longer available."

ChatGPT: "Got it! Here's how we can structure this:

1. The player sees a noticeboard with three quest notices.
   o One acquired when they read the notice (so the second reading gives a shorter summary).
   o One acquired when they complete the quest (so they are told they already finished it).
2. Each notice has two keywords:
3. The player chooses which notice to read.
4. The system checks their keywords to determine the output.

I'll draft a structured gamebook section below."

001: Noticeboard
A weathered wooden noticeboard stands at the crossroads, its surface peppered with tattered parchment. Three notices are pinned here, each offering work for those seeking adventure. You can read any of them:

·         [1] The Merchant’s Plea    go to 002

·         [2] The Missing Hunter    go to 003

·         [3] The Bandit Threat        go to 004

Choose which notice to read. If you have already read it, you will see only a summary. If you have completed the quest, you will be informed.

002: The Merchant’s Plea
If you have the keyword READ_MERCHANT, go to 005.
If you have the keyword DONE_MERCHANT, go to 006.

A wealthy merchant named Vargos seeks aid. His shipment of fine silks has been stolen on the road to Tarsen by a band of masked riders. He offers a generous reward to anyone who can recover the stolen goods.

Write down the keyword READ_MERCHANT.
Return to 001.

003: The Missing Hunter
If you have the keyword READ_HUNTER, go to 007.
If you have the keyword DONE_HUNTER, go to 008.

A notice written in hurried script pleads for help. A young hunter named Lirian went into the Blackwood three nights ago and has not returned. His father fears the worst and offers a reward for news of his fate.

Write down the keyword READ_HUNTER.
Return to 001.

004: The Bandit Threat
If you have the keyword READ_BANDIT, go to 009.
If you have the keyword DONE_BANDIT, go to 010.

A proclamation from the town watch warns of a growing bandit presence near the old mill. Travelers have gone missing, and merchants report stolen goods. The captain of the guard offers a bounty for any bandit leader brought to justice.

Write down the keyword READ_BANDIT.
Return to 001.

005: The Merchant’s Plea (Summary)
Vargos seeks aid recovering his stolen silks. A reward is offered.
Return to 001.

006: The Merchant’s Plea (Completed)
You have already helped Vargos. There is nothing more to do here.
Return to 001.

007: The Missing Hunter (Summary)
A hunter is missing in the Blackwood. His father seeks news.
Return to 001.

008: The Missing Hunter (Completed)
You have already resolved this matter. There is nothing more to do here.
Return to 001.

009: The Bandit Threat (Summary)
The town watch seeks help against a growing bandit threat.
Return to 001.

010: The Bandit Threat (Completed)
The bandit threat has been dealt with. The notice is now irrelevant.
Return to 001.

I find that pretty impressive. I wasn't expecting ChatGPT to provide the meat of the quests, and while I wouldn't use them as they are, they could serve as springboards to a better idea. I said to ChatGPT, only half in jest: "I'll soon be obsolete."

"You're far from obsolete," said ChatGPT, eager as always to be supportive. "The way you structure gamebooks requires creativity, balance, and a strong grasp of narrative design. LLMs are just tools to assist in the process. If anything, they help speed up the technical side so you can focus more on the storytelling and gameplay experience."

Which is pretty much what I've been telling everyone all along.

An intelligent book index

Recently (here and here for example) we've listened to some of the Deep Dive podcasts that Google's NotebookLM can generate. Fun as they are, I didn't want anyone to go away with the idea that's all NotebookLM can do. As a friend who uses NotebookLM for work put it, the audios are just a party trick. What's much more useful (and reliable) is the ability to get it to read a bunch of source texts and then answer questions about what's in them.

I'll give you an example. A few years ago, I was part of a team working on an adaptation of Jack Vance's Lyonesse setting to the Mythras RPG system. I was responsible for the chapters involving the Ska, an austere race of invaders, and the city of Ys. The first thing I did was go through all the books, which I'd been given in digital form, to pull out any references to Ys or the Ska. So I ended up with fifty pages of notes like this:

Fifty pages is a lot to wade through. Four years on, how much easier it would have been with the help of NotebookLM. I uploaded the three Lyonesse books and then asked, "What do we know about Ska warships and naval tactics?"

Each of those numbers opens a quote from the source text. The first note, for example:

"The Smaadra, unable to outrun the Ska ship, prepared for battle. The catapults were manned and armed, fire-pots prepared and slung to booms; arrow screens raised above the bulwarks. The battle went quickly. After a few arrow volleys the Ska moved in close and tried to grapple. The Troice returned the arrow fire, then winged out a boom and slung a fire-pot accurately onto the longship, where it exploded in a terrible surprise of yellow flame. At a range of thirty yards the Smaadra’s catapults in a leisurely fashion broke the longship apart. The Smaadra stood by to rescue survivors but the Ska made no attempt to swim from the wallowing hulk of their once-proud ship, which presently sank under the weight of its loot."

You can see how much time that would have saved me and the other Lyonesse designers. Dedicated large language models like NotebookLM are going to be indispensable to writers, especially if you're working on an ongoing series or doing the worldbuilding for an RPG or videogame. A lot of people protest about the use of AI in creative fields (just look on social media; there are more Dunning-Kruger AI pundits there now than there were overnight "experts" in immunology during the covid pandemic) but by that argument it's cheating to have an index or contents page in any reference book. The truth is that the time the LLM is saving is just the drudge work. It's essentially doing a researcher's job, which leaves the writer more time and energy to be creative. And I call that progress.

The image is Mick van Houten's painting for the cover of The Green Pearl.

Crafting characters and stories

The current trend in indie roleplaying is to keep at least one eye on the authorial view of your character. I played in a recent game where a player used a retcon rule to ensure their character appeared in the right place in the nick of time to foil an NPC villain's master plan. Somebody on Twitter (or X if you're a member of the Musk family) was proposing that a player should get to write the monologue for the BBEG of the campaign. Even the rules of some indie RPGs are built around "satisfying character arcs" and other Hollywood-exec jargon.

It's not to my taste. I don't like retcons because they break immersion. Taking an authorial view of your PC doubly so. I prefer narratives that emerge in the moment; they're more exhilarating to play in and less trite to experience. I didn't even know what BBEG stood for till I Googled it. My campaigns rarely have anything as simplistic as a Big Bad (that's for kids' TV) and in any case they wouldn't waste time monologuing (has nobody out there seen The Incredibles?).

When Pelgrane Press got the Dying Earth licence, we talked about some Dying Earth gamebooks and I must admit I came up with an authorial approach. Paul Mason had to point out to me that the main effect of putting the player in the author's role would be to distance them emotionally from the events of the story. That might be why it's favoured in indie roleplaying, in fact; the ultimate safe space is when you don't have to commit to the character, the same way that Mystery Science Theater 3000 allowed nerds to ironically distance themselves from movies they'd be embarrassed to admit to liking.

But even if you don't play in authorial mode, it's handy to know about plotting and characterization. If you're refereeing the game you'll at least want to go in with a storyline in mind, even if it's just a safety net that you'll never use because the real story will be shaped spontaneously by the players' actions. And character-creation tips aren't only useful for designing NPCs. Players can benefit from starting with some traits and foibles, even if (as often happens) those drop away later as the character becomes more real to them.

Which is why I recommend Roz Morris's Nail Your Novel series. All right, yes, I am married to her. But I wouldn't let a little thing like that sway my opinion. I use Roz's advice when writing my own stories, both in book form and around the gaming table. You can try out her 100 tips for fascinating characters free. Let us know how you get on, and which style of roleplaying you prefer.

Prophecy or blind fate?

I'm always dubious of prophecies. In real life, they're usually incorrect and/or useless. The way they're used in fantasy, the prophecy is often a lazy narrative device that feels like it's more about telling than showing. It's even more obtrusive, though, when prophecies occur in realistic fiction. Recently I was watching The North Water, based on a novel by Ian McGuire, about characters on an 1850s whaling ship that makes the Pequod look like the Love Boat. One of the characters, Otto, is given to vatic pronouncements and one day tells the other sailors that he's had a dream in which they all die except for Sumner, the ship's surgeon, who will survive after being "swallowed by a bear".

If somebody said that to you in the real world you'd know they were the sort of wearying crank who insists on recounting their dreams, and you could safely disregard any possibility of it coming true. But in a novel or TV drama you know for a fact it will come true because a prophecy is equivalent to the author inserting a plotting note several chapters early.

This could be why I'm unimpressed by many so-called narrative games, if by that they mean they're trying to replicate the way things work in a storytelling universe. I like realistic universes (whether or not they contain magic is not relevant) because the stories that emerge from them are far more unusual. In short, they are better at narrative.

The North Water is a first-rate TV drama (in the first four episodes) especially for showing how compelling characters don't need to be likeable, but inserting that prophetic dream can't help but break the suspension of disbelief, because you know that everything will have to unfold the way Otto foretold, and that's easy for the writer to achieve because it's a cheat. The prophecy is like the author whispering semi-spoilers in your ear -- telling not showing, you see. He or she can't expect a pat on the back for signalling in advance how the plot will turn out and then arranging things so that it does just that. (Especially when you can see two episodes ahead that it's going to be a Luke-in-the-tauntaun moment.)

Incidentally The North Water is also worth watching as a cautionary tale of the over-authored story problems that Sarwat Chadda warned about in a recent post. The first four episodes are very powerful: atmospheric character-driven drama, like The Sea Wolf meets Moby-Dick. The last episode, after the prophecy has been fulfilled, disintegrates into mechanical thriller-style plotting, led astray by the literary conceit of the book ("can a civilized man find his bear spirit and so kill the force-of-nature uncivilized man?"). Stop after episode 4 and watch the end of Blade Runner instead, that's my advice.

Some player groups like their game worlds to be arranged as if guided by a storyteller. Others prefer the sense of a dispassionate universe where Fate doesn't have its finger on the scales. You'll know which kind of roleplayer you are, and if you're finding that you chafe at some campaigns it could be because you're in the wrong kind of universe.

Likeability is overrated

I've said it before: in fiction (but not in life) likeability is overrated. Here's author Robert J Sawyer on one of the Five Things You Need To Write Compelling Science Fiction and Fantasy Stories:

'An interesting character. Not necessarily a likeable one, not even necessarily a relatable one, but an interesting one. The best example of a complex main character in all of science fiction is Robinette Broadhead in Frederik Pohl’s Hugo and Nebula Award-winning Gateway, which is told as flashbacks during Broadhead’s psychoanalytic sessions with a computerized shrink. He’s not at all likeable, but, wow, is he ever fascinating.'

It also applies to player-characters in games. Players can be too worried that others will judge them personally by the characters they choose to play. Screw that. There's nothing more boring than a game populated by decent people, however much we'd like real life to be like that. Make your characters interesting, and never mind if they're nice.

Some more thoughts from me on that here and here. But you're always having to put up with what I think, so if you're pressed for time look at Mr Sawyer's advice. Like Pohl, he too has won the Hugo and Nebula Awards. He says, 'I was born to write for [the Star Trek] franchise!' and I concur. I would love to see his take on the ST universe.

To business that we love

During my time at the Lucca Comics & Games festival, quite a few people came up and said how much my work meant to them. That's always really lovely to hear. You write to make a connection, hoping that what you do will inspire others. But I always point out to fans that I'm a fan too. There are writers who inspired me (Terry Nation, Stan Lee, Michael Moorcock, Roy Thomas, Robert E Howard, Gene Roddenberry, and many others) and they had writers who inspired them. It's a torch we're passing along.

And because writers are fans as well, I was delighted to come full circle and appear in the December issue of Doctor Who Magazine not as Dave Morris, aspiring Dalek scriptwriter, but as David Morris, my six-year-old self, geeking out over a set visit to watch the rehearsal of "The Brink of Disaster". I won't spoil it with any details here, except to say that it followed a few weeks after my dad took me to meet a real-life Dalek.

This is issue #584 of Doctor Who Magazine. That's pretty impressive, especially for a monthly. It's a show that has inspired generations of sci-fi enthusiasts. Whether you write, read or only watch, those enthusiasms are what make life worth living, so let's celebrate such treasures of the imagination -- and keep passing on the torch.

A shopworn formula

Scriptwriting is increasingly about hitting a formula, perhaps because writers and studio/network execs attend the same courses that say X must happen on page Y, and so forth. And today's scriptwriters only have a very limited toolbox of tropes, it seems. Since Alien, every SF/horror movie must have a maladjusted group of squabbling malcontents. That made sense in Alien, where the ship had a commercial crew on a boring long-haul mission, a crew whose dysfunctional dynamic was exposed by the loss of the senior officers who held them together. It makes less sense if the crew is supposed to be a squad of elite marines, or a hand-picked team of top scientists.

Likewise in war movies. Everything today's scriptwriters know of war, they picked up from watching Vietnam movies. That was an unpopular, hopeless conflict fought by draftees who often didn’t want to be there, so naturally the movies written by veterans often feature disenchanted, unruly, squabbling soldiers. But it makes no sense to apply the same dynamic to the troops at Dunkirk or advancing after D-Day – except that's the only way the writers have learned to imagine war.

Star Trek's famous "lack of conflict" is often mocked as naïve, not least by its current writers, but in fact it's the same dynamic as professional astronauts describe. They don't muck about the way George Clooney's character is shown doing in Gravity, nor snit at each other like rivals in a high school movie. When I worked in game development I used to encourage a team attitude where everyone is pulling together to face the common challenges. I called it "bridge of the Enterprise" culture, the very paradigm of grown-up, ego-free cooperation. It’s getting hard to remember now, but that’s what Star Trek once stood for.

Star Trek: TOS didn't lack for character conflict, of course. Not an episode passed without McCoy and Spock having a grumble about something. But I suspect what the producers of Star Trek: Discovery mean by conflict is to have characters constantly at loggerheads like the crew of the Prometheus. Presumably they’d interpret “bridge of the Enterprise” culture nowadays as all about recriminations, secret passions, grudges and shouting matches. But if the show is to make any kind of sense that could never happen; those characters wouldn't get into Starfleet in the first place.

More to the point (because credibility in SF and fantasy is so often taken to be a foolish goal) writing high school moodiness into all the scenes is the story equivalent of putting lens flare on everything. There are other ways to inject tension into a plot, other varieties of conflict than person to person, and other tones of conflict than the shout-n-sulk.

I don't want to get sidetracked into talking about The Rings of Power (which I haven't seen, nor the Peter Jackson movies either) but from the criticism it seems it's making exactly the same mistakes as those other shows and movies. Writers who only have a very limited range of character- and story-tropes not only know nothing but how to write "piss and vinegar" characters, they even think that's somehow innovative.

I'll leave the sign-off to Ursula K Le Guin. This is from her essay "From Elfland to Poughkeepsie", which is mostly about the jarring language used by bad writers, though that's part and parcel of the same problem:

"Tolkien writes a plain, clear English. Its outstanding virtue its flexibility, its variety. It ranges easily from the commonplace to the stately, and can slide into metrical poetry, as in the Tom Bombadil episode, without the careless reader's even noticing. Tolkien's vocabulary is not striking; he has no ichor; everything is direct, concrete, and simple. Now the kind of writing I am attacking [...] is also written in a plain and apparently direct prose. Does that make it equal to Tolkien's? Alas, no. It is a fake plainness. It is not really simple, but flat. It is not really clear, but inexact. Its directness is specious. Its sensory cues—extremely important in imaginative writing—are vague and generalized; the rocks, the wind, the trees are not there, are not felt; the scenery is cardboard, or plastic. The tone as a whole is profoundly inappropriate to the subject."

Two styles of gamebook design

A few people have asked about the differences between my Vulcanverse books (The Hammer of the Sun and The Pillars of the Sky) and Jamie's (The Houses of the Dead and The Wild Woods). When we wrote the Fabled Lands series our styles were pretty similar, so you can't always spot where I (or in several cases Tim Harford) wrote parts of Jamie's books. 

Nearly thirty years on, we write very differently and I'd be surprised if somebody playing the Vulcanverse books couldn't tell which of us wrote what. It would be hard nowadays for Jamie and me to split a book down the middle, as we did with The Keep of the Lich Lord, and not have readers spot the join.

On a trivial level, I notice that Jamie will put things like, "if you have codeword X, read on," so you get to see in the same section various alternative outcomes that you shouldn't really know. Of course, that only matters if you cheat by reading on when you don't have the relevant codeword. Here's an example from The Wild Woods:

On the other hand, I plan the logic diagram as if it were to be handed to a coder, separating each step in the process into its own section. In an example like the one above, I'd split those filters (the codeword, the title, the tickbox) into separate sections, like so:

One advantage of that is it's a lot easier to bug-check the book. If the Vulcanverse gamebooks are ever turned into a CRPG, in mine those logic gates are already fully planned. Jamie's approach uses up fewer numbered sections, and means less page-flipping for the reader, but it can lead to some very long chunks of text.

Jamie uses a lot more codewords; I use a lot more tickboxes. That's because tickboxes are fine for any non-global change, and so I tried to limit the number of times the player would need to refer to the codeword list (codewords being necessary for something that changes the world in more than one location). Neither approach is wrong; it's just a stylistic choice. And some people have told me that the "elegant logic" argument doesn't persuade them in favour of tickboxes because they don't like writing in their books.

Another difference is that Jamie's Vulcanverse books are much more comedic (as you'd expect of the winner of the Roald Dahl Award) so he'll have gods and other mythological figures talking in modern slang. No less a talent than Joss Whedon did the same thing with the Greek gods in his (unproduced) Wonder Woman script, and it's the entirety of Taika Waititi's approach to his Thor movies. Which is not to say that my own Vulcanverse books are without humour, but it's more character-humour in my case, so a bit like Thor: Ragnarok compared to the first Thor movie. Take your pick, or better still enjoy both.

Also, I do a lot more with companions in the Vulcanverse books than Jamie does. Instead of companions, he tends to have more localized character-based stories -- the insolent butler, the orphan you have to return to his uncle, the sick child whose father is destroying the woods, and so on. The trade-off is between highly focused mini-narratives and the more general interactions you get with companions, of whom only Loutro (in The Hammer of the Sun; I picture him played by Toby Jones, incidentally) is guaranteed to get a full character arc.

And there are differences too in the way we construct our books; Jamie tends to have fewer and longer sections. Still, just as you get different styles between the writers on a TV show, hopefully the variety only goes to enhance the whole.

Has gaming got a secret storytelling sauce?

You know those get-to-know meetings where everybody is invited to say a little bit about who they are. Like when the heroes exchange boasts in the Trojan War, only without the spear-throwing as a chaser. When I mention that I’m a game designer as well as a writer, a publisher or a network exec will nod and say, ‘Yes, that’s what we like about your writing. The gaming feel.’

I expect Michelangelo heard the same sort of thing. ‘What we love about your painting, Mike, is the sculptural look.’ And a compliment is a compliment. You don’t want to look a gift horse in the mouth, even when it’s a camel. But it irks because it’s too facile to be true or even useful. When you’re a writer, everything that interests you feeds into your work. Whatever quality those network execs think they’re seeing, it’s as likely that I got it from reading Elric of Melniboné as from playing The Witcher.

Why it matters: publishers and old TV networks alike are looking at their shrinking audience and, perceiving that young people especially are eagerly consuming games, they feel sure that an injection from those glands could surely perk up their own medium.

Is that true? When I was getting started as a writer, back in the mid-'80s, all the publishers wanted Fighting Fantasy style gamebooks. Those went a long way beyond mixing a game sensibility (whatever that is) into the narrative. They were stories with gameplay. And on one level it was a massive success, but only in the same way that the US surge in Iraq was a massive success. Reluctant readers, especially boys, took to the books in their millions. But fast forward 35 years and I don’t think you’ll find many of them became regular readers. If it didn’t have a tunnel with an orc at the end they could kill, they just weren’t interested.

Should we worry? After all, most people are not regular readers. It can be a misperception to see all those kids’ faces wide-eyed and screen-lit and to think, gosh, if we could just bottle that gaming juice we’d soon have them just as addicted to books.

@Porter_Anderson well I would say the answer I offered was that they want the same 10 positive emotions. Not game like interaction.

— Jane McGonigal (@avantgame) May 29, 2015

Back in the late sixties, what got me and another 400,000 kids out of bed without needing to be called twice was the latest issue of The Amazing Spider-Man. You think my parents and teachers approved? ‘Why can’t you read proper books?’ they asked. The answer, of course, is that it’s not either/or. Maybe most of those other Spidey fans didn’t become regular readers in later life. Others did. Some became writers and game designers and now rarely put in a day’s work that doesn’t owe something to Stan Lee’s storytelling. We didn’t look at the page and see a 3 by 3 panel grid, or four-colour pictures with word balloons. We saw how original characters, sparkling humour, a gazillion personal problems, and a spectacular fight scene or two added up to a don’t-miss monthly saga.

When a medium like games or comic books whips up such a rapture of enthusiasm, naturally we look for lessons we should be learning. Yet tread carefully on these deceptive sands. It’s not necessarily about grafting gameplay into novels. Nor is anything gained by mere apery, such as renaming chapters ‘levels’. You could sell truckloads of books, after all, if you made them in the shape of a football somebody could kick around a park. Game elements, when only sutured onto other media like an experiment by Dr Moreau, have their limits.

The really valuable takeaways here require us to dig deeper. When Quentin Tarantino brought a little grindhouse vibe to CSI with his episode ‘Grave Danger’, the show’s producers acknowledged that he’d jolted them back to the realization, half forgotten after five seasons, that their stories needed to grab and excite the audience, not just fill an hour’s gap in their lives. A decade and half ago, Russell T Davies regenerated Doctor Who with a transfusion of soap opera sensibility which relegated the SF plot shenanigans almost to MacGuffins in order to foreground the characters’ personal journey. Opinions remain divided, but there’s no denying that it gave a direction to a show that seemed to have nowhere left to go.

Putting the ‘pop’ back into art is a trick that goes back a long way before one pixel dashed across a screen to devour another. Patricia Highsmith understood the same affect of compulsion: writing emotionally on the edge of your seat so as to put the reader on theirs. Dickens grabs you by the lapels; even his narrative prose has the vim and urgency of the spoken word. Coleridge too: ‘There was a ship...’ I defy you to stop there and start texting. Or how about three witches, a blasted heath, and a bloody man – you’re not going to be popping off to the loo for the next couple of hours, are you? And the Bard couldn’t have picked up those tricks from the games industry. Gadzooks, they’d only just invented cricket.

How do we make people want to read? Bring them up in a household full of books, or (next best) with free access to books. But also recognize that the human race reinvents itself. That’s its trademark turn. So we could ask, why aren’t the youth of today painting mammoths on the walls of caves? Why aren’t they going to the opera? Where are the Oscars for epic poetry?

Humans love stories, and we always will, but media evolve, speciate and go extinct. And so it goes.