From The Iliad

Vaunt Lament

To the man I have mortally wounded, as I strip him naked

Never mind your pride. Forget how you never grow too tired to fight. 

Even so. You show how you Torys shall quit the ships of the Achæans, 

by every disgrace with which you’ve ever tried to tar us described. 

Wolfbitch cowards yet you wouldn't cringe and follow Zeus, would you?

Not Zeus who punishes bad guests and one day's going to butcher your city.

When you stayed with me you took my wife, you stole such a lot of treasures,

after all she did to make you feel at home,

now you want our ships on fire and to kill us all.

Still there will come a day on which, despite your zeal for Ares,

you’ll be stopped. Oh father Zeus, you are the wisest

of every god and human, this is all your fault.

You favor them I kill, why, so boastful and bloodthirsty?

After a long spell, a person can tire of anything at all.

Of sleep, love, song, dancing, still these are things whereof 

you’d always want more than you’d want of war.

Yet they never tire of wanting more war.

Kampala Yénkya

Five short stories by Dilman Dila, which served as inspiration for the Kampala Yénkya climate futures TTRPG we worked on together.

AI Literacies and Media Education

 A sort of mini-roundtable, thrown together quickly in response to ChatGPT and other generative AIs: AI Literacies and Media Education.

Utopian Heists

From a glossary-in-progress.

Utopia may conjure up the notion of a perfect society, or perhaps an apparently perfect society with a dark secret (so actually a dystopia, or anti-utopia). Within utopian thinking and practice, however, utopia is a much more flexible term. Sometimes it refers to communities that attempt a better way of life, or plans for such communities, or stories about such communities. But often it refers to something more like a method, a way of thinking about the world around us, and perhaps about how we live within its constraints, and what we can change in the short term as well as the long term.

Heist scenarios in TTRPGs involve players working together to plan and execute a high-stakes theft or robbery. Players will typically gather information, develop strategies to overcome obstacles, execute their plans, and then (when nothing goes according to plan) improvise. Typically each member of the caper crew is both exceptionally skilled in their field, and also a kind of eccentric and maverick; for Fredric Jameson, this is how the heist plot becomes a ‘distorted expression of the utopian impulse insofar as it realizes a fantasy of non-alienated collective work.’

There are plenty of variations on the heist model, including smuggling, spying, scouting, seige-breaking, hijacking, sabotage, kidnaps, rescues, jailbreaks and exfiltrations. A ‘reverse-heist’ scenario involves sneaking something into a secure location. Often, as in Christopher Nolan’s film Inception (2010), it’s not enough to get in and out again safely, since the reverse-heist will fail if anyone finds out that the package has been planted. A red / blue team exercise is a sort of officially sanctioned fake heist (see Red / Blue Team Exercise). We might think of critical utopias, in the tradition of Ursula K. LeGuin’s novel The Dispossessed (1974), as a kind of prefigurative red-teaming, seeking to improve the robustness of a system that has not yet even been built (see Critical Utopia).

One influential TTRPG centred on heists is John Harper’s Blades in the Dark (2015), where players are a crew of daring scoundrels seeking their fortune in a gritty, industrial-fantasy city. Many other heist-ready games spring from the shadows of its Forged in the Dark system; examples include John Leboeuf-Little and Stras Acimovic’s Scum and Villiany (2018), Kienna Shaw and Jamila R. Nedjadi's Songs for the Dusk (2021), and Andrew Gillis’s Girl by Moonlight (2024). Grant Howett’s Honey Heist (2017), at just one page, is another influential heist TTRPG, in which the players portray bears. Howett is fond of capers: in Royal Blood (2020), you effectively play a living Tarot card on a heist-like mission; One Last Job (2014) assembles a crew of washed-up criminals (or “WW2 commandos, ageing punk rockers, zombie apocalypse survivors, Wild West cowboys, and so on” (p. 3)), with character creation embedded in gameplay through improvised reminisces and banter. Jason Morningstar's Fiasco (2009) is another fiction-first game that lends itself to bungled capers of all kinds, and some of its playbooks are specifically heist-themed. Beyond the TTRPGs that revolve around heists, a heist episode may turn up in any campaign. San Jenaro Co-Op’s The Roleplayer’s Guide to Heists (2019), for example, an anthology of system-agnostic heist scenarios.

In fact, heists have a lot in common with a classic staple: the dungeoncrawl. Both dungeoncrawls and heists are a kind of breaking-and-entering. Both revolve around the central theme of embarking on a mission filled with mystery and peril, where the primary goal is to obtain something of value. They often feature complex environments with many opportunities for players to work as a multidisciplinary team of specialists, formulating and enacting cunning strategies. We might even imagine a kind of spectrum, or some other visualisation, of these two primary modes of tabletop thieving.

So what is the difference? Well, we might say that crawls place more emphasis on combat, and that heists are about stealing something in specific, rather than accumulating whatever treasure is lying around. Then again, heists frequently include or degenerate into bloodshed; moreover, dungeoncrawlers may well have a final treasure in mind (the Amulet of Yendor in the digital game Nethack, for example).

Perhaps crawls have some affinity with epic fantasy settings, and heists with modern urban settings. Heists may involve a ticking clock, whereas crawls are more relaxed — ‘relaxed’ in the sense of dragging your bloodied bodies back to town for a long rest before attempting the next raid. Heist crews are often assembled by a mastermind, and might not have previously known each other (let alone trusted each other) — dungeon adventurers are often drinking buddies. Perhaps the dungeoncrawl also has a certain ideological tie with colonial conquest, which the heist does not: some dungeoncrawls are genocidal. But none of these are hard-and-fast distinctions either.

One difference is between the intricately interlinked challenges predicted in the heist plan, vs. the diverse and unknown dangers of the dungeon. This is reflected in subtly different kinds of division of labour. The heist and the crawl both demand a mix of specialists, of course. But the trope that only so-and-so can perform vital task x is more characteristic of the heist. Heist crews are made up of supremely respected workers. Should any one of them withdraw their labour (perhaps even for a few seconds), it could send out shockwaves like a general strike. In ‘A Global Neuromancer’ (2015), Fredric Jameson contrasts the familiar picture of the division of labour (à la Adam Smith) with heist-divided labour:

[...] specializations are certainly present—we need someone to open safes, someone acrobatic enough to get through windows, someone capable of neutralizing the alarm system, someone to drive the car, someone to secure the plans on what is probably going to be an inside job, and finally the brains or the mastermind, who is also the political leader so to speak. But each of these characters will be idiosyncratic: it is a collection of interesting oddballs and misfits, all of them different, and many of them in serious personality conflict with each other. The technological features of the object have thus been humanized and personified if not altogether sublimated [...]

Jameson suggests that heist plots are therefore allegories of production (and he adds, a little mysteriously, that they are allegories of the inner divisions of the psyche too). Why then are there not more heist-like narratives about making things? Perhaps because it would be too on-the-nose. Jameson also offers another reason.

Here I think we face the dilemma of any literary or artistic representation of labor: it is very rare indeed that the content of the industrial product can have any necessity. The production process itself is always interesting: but [...] the nature of the object cannot have any real aesthetic necessity without turning into a symbol of some kind.

Why can’t the nature of the object have any real aesthetic necessity, except by turning it into a symbol of some kind? Because whether the fictional characters are creating chamber pots, or machine parts, or sausages, we know that “[a]nything that can produce a profit is equivalent when it comes to generating surplus value,” i.e. what they are really making is money.

This is a very Jamesonian preoccupation with large-scale material conditions constraining, and being reflected in, what artworks can and can’t do. Jameson does think that the heist plot gives it to us straight. It does this either by making the macguffin very macguffiny, or by simply making it a large sum of money. The score needs to be divisible, after all — the heist team wouldn’t share. In this way, for Jameson, the heist genre “short-circuits the search for a meaningful object by simply positing the cash, the gold, the bearer bonds, or whatever else. So this is, as it were, the negative or critical, the demystifying side of the caper form.”

There is a related factor that also distinguishes the heist from the crawl. It’s a controversial word within utopian discourse: planning. A heist is all about meticulous planning. The thrill of a may can come from improvisation as these flawless plans fall apart. But it can also come from things going off without a hitch. By contrast, the dungeoncrawlers just equip themselves as best they can, and venture into the unknown. They may even invent some mini-heists when they’re down there, but they haven’t spent weeks trapped in a planning montage.

It’s worth dwelling on planning for a bit. The word means a lot of things. In architecture, “planning” pertains to designing physical spaces to meet both functional and aesthetic requirements. With respect to political economy, “planning” might mean defining clear parameters for the production and distribution of goods and services. Broadly speaking, in a market economy, the interaction of economic supply and demand determines what goods and services are produced, and who gets to benefit from them; in a planned economy, these decisions are made by the government (if you want to make it sound sinister) or by the people (if you want to make it sound optimistic). The socialist calculation debate, which emerged in the 1920s, stemmed in essence from the attempts of various neoliberals (such as Friedrich Hayek) to prove from first principles that a complex modern economy is unplannable, and that any planned economy is a contradiction in terms and a catastrophe waiting to happen. Of course, the crude distinction between “market economies” and “planned economies” does not reflect the complexities of real-world economies, and is also heavily shaped by the ideological presuppositions of mainstream economics. Taking a more nuanced and interdisciplinary perspective, market economies also include plenty of planning, and planned economies have many elements of distributed decision-making.

For example, governments play a fundamental role in creating, maintaining, and shaping market mechanisms and forces. They establish the legal and regulatory frameworks that define the ‘rules of the game’ within which markets operate. These cover areas such as banking and finance, employment law, property law, contract law, competition policy, and sector-specific regulations. Governments also maintain markets through regulatory oversight and interventions to ‘correct market failures,’ such as monopolies, information asymmetries. They run themselves ragged trying to internalise externalities, including the greenhouse gas pollution driving climate change. Furthermore, governments shape market forces through their policy decisions, e.g. fiscal policy (taxing and spending) and monetary policy (e.g. interest rate adjustments by central banks).

Likewise, even if we take the example of the “material balances” planning of the early Soviet Union — as close as you’ll get to a textbook planned economy — what the central agency Gosplan did, strictly speaking, was to calculate interactions of supply and demand in order to produce and distribute goods and services. It is not normally described in these terms, but it should be. Yes, these determinations of supply and demand via demographic data, statistical calculations, political priorising, and other non-monetary means, were often wildly out of touch with people’s real needs, desires, capacities, risks, and so on. But the same can absolutely be said of supply and demand determined by financial allocations in a market economy.

Then there is “planning” at the organisational level, which will typically be differentiated from “strategy,” “governance,” “management,” “risk management,” as well as future-oriented processes like “horizon-scanning,” “anticipatory governance,” “Research & Development,” “stakeholder engagement,” “investor relations,” “cost-benefit analysis,” and so on. It will also be differentiated from vaguer terms like “commitments,” “promises,” “vision,” “ambitions,” “scoping,” and so on, whose meanings tend to be less specified by legal and regulatory practices. Planning tends to imply relatively detailed thinking about implementation of pre-given goals; it might be thought to sit “below” strategy and governance but “above” more day-to-day management. Nonetheless, despite effort at differentiation, planning overlaps with all of these. Perhaps what is most striking, at the organisational level, is the the division of anticipatory labour. Does this fragmentation of anticipation imply that, just like a party of adventurers can achieve more than the sum of its parts, an organisation is able to imagine possible futures more boldly and precisely than it otherwise would? Or, in line with the old adage “never split the party,” is such fragmented anticipation more restricted in the futures it can envision and steer towards? Probably both things are true, in different ways. There are also intriguing questions around what a different division of antcipatory labour might accomplish.

Then there is urban planning, which is particularly entwined with utopian discourse (along with architecture). Urban planning is inherently transdisciplinary, drawing on many strands within public policy and social science, and in this sense may have some affinity with the sometimes totalising aspirations of utopian thinking. Urban planners may consider issues of sustainability, development, health, economic growth, quality of life, crime and policing, social cohesion, and so on. But these are clearly all contentious terms — and some contemporary planners may well be interested in postgrowth rather than growth, or postdevelopment rather than sustainability, or police abolition rather than crime and policing. The history of urban planning is also the history of a variety of practical utopian experiments. For example, in the postwar period in the United Kingdom, the ‘new town’ movement embodied a similar mix of energies to the emerging welfare state. According to Rosemary Wakeman, these new towns, garden cities, cities of science, etc., were

both a reflection on and a critique of mid- to late twentieth-century society. A steadfast belief in physical determinism was shared across the architectural and planning professions. An ideal social atmosphere could be achieved by carefully planning all the physical elements of the city. Designing the physical fabric would change individual behavior, social relations, civic life, and community. The assumption was that the ideal city could be mass-produced for a mass cultural age. Life would be balanced and harmonious.

(5)

To this rudimentary survey of planning discourse — planned economies, planning within organisations, urban planning — we can add heists and dungeoncrawls. Heists and dungeoncrawls make implicit claims about what can legitimately be planned for and what cannot, about how to do so, and about how and why the best-laid plans often go awry.

The degree and type of organisation within the target stronghold is also worth mentioning. Heist and caper crews, despite their superlative skills, are often framed as underdogs. The players’ power comes from stealth, deception, and cunning: they are unlikely to succeed at blasting or hacking their way into that stronghold. But a party at the dark mouth of the dungeon stands on another kind of threshold. True, they are also underdogs, in the sense that the army of monsters massed down there could probably easily overwhelm them. Except that it is not really an army. Unlike the guards in the stronghold, the dungeon denizens are typically fragmented. They may be antagonistic or indifferent to one-another, and their interests and activities do not constitute a joined-up defense of the territory. Perhaps to the extent that the dungeon does mount a dynamic, coordinated defense, the crawl becomes less like a crawl and more like a heist, raid, or siege.

Occasionally the desire to serve up the dungeon in digestible chunks may become an absurdity, something we are expected to suspend our disbelief about, as a convention of the genre. Monsters may even chivalrously attack one at a time (“It’s called class, Rick, it’s called class,” as one self-aware comic puts it. TTRPGs in which dungeon dwellers acquire a kind of class consciousness could be interesting to explore). Many TTRPGs, such as Grant Howitt’s TTRPG Goblin Quest (2015), as well as Grunts! (1992) by Mary Gentle, Dungeon Crawl Inc. (2021) by Dakota Krout, Dungeon Keeper Ami (2005-ongoing) by Pusakuronu, Ralph Breaks the Internet (2018), The Order of the Stick (2003-ongoing) by Rich Burlew, Henchgirl (2015-2017) by Kristen Gudsnuk, etc., explore the inner lives of mooks and minions, sometimes highlighting these convention.

A few other distinguishing features are worth considering. Heists often involve an insider; a less frequent trope for a dungeoncrawl, although one with considerable potential.

Another common feature of heist narratives is the twist: the double-cross, the mole, the apparent slip-up that was actually part of the plan all along. Maybe the thing we thought we were stealing is really something else, or never existed in the first place. The improvisational nature of TTRPGs might present challenges for such storytelling, insofar as a truly impressive twist is typically carefully prefigured. Of course, the significance of past events can always be retrospectively altered, in service of a twist. Moreover, a Blades in the Dark mechanic allows most of the heist planning to take place ‘off screen’, and to be revealed when it is narratively salient via a flashback:

A flashback isn’t time travel. It can’t “undo” something that just occurred in the present moment. For instance, if an Inspector confronts you about recent thefts of occult artifacts when you’re at the Lady’s party, you can’t call for a flashback to assassinate the Inspector the night before. She’s here now, questioning you—that’s established in the fiction. You can call for a flashback to show that you intentionally tipped off the inspector so she would confront you at the party—so you could use that opportunity to impress the Lady with your aplomb and daring.

The Rick and Morty episode ‘One Crew over the Crewcoo's Morty’ (2019) satirises contemporary Hollywood heist cinema with a Heistcon, a heist-off, and the trivial automation of the supposed novelty and surprise of heist-style twists and turns. It plays with various ways of extrapolating heist norms beyond their usual scope. One scene imagines what might happen if everyone (hundreds or thousands of people) were part of a heist crew. Another scene is a long back-and-forth between the Rick and the Heistotron machine devised to automate heist design and delivery. This scene lampoons the imaginative poverty of at least one kind of heist ‘twist’: I knew that you would do A, so I did B, except you knew that I would do B, so you did C, except I knew that you knew that I would do B and you would do C, so I did D, and so on. There is a certain psychological unbelievability to this kind of twist. It takes a style of thinking characteristic of ruminative, anxious obsession, and imagines it having purchase on social reality. We also might detect a kind of neoclassical economic logic, one which emphasises the utility-maximising behaviours of individuals given the the information available to them, behaviours conceived of as relatively independent of institutions, norms, cultural practices, and so on. The messiness and uncertainty of how others desire and act is removed, and all it takes to understand what someone will do is to have enough information about the information they have access to. Yet I knew that you knew that I knew is also a kind of quasi-erotic fantasy of cognitive intimacy, albeit antagonistic, which resonates with utopian themes, and stands in some tension with such neoclassical logic.

Dystopian societies, including fake utopias, are obvious targets for heists. But what about other types? Is utopia ripe for a heist? What is there to steal? Clearly it depends on the utopia. Many utopias have abolished money, or perceive value according to some novel scheme. In Voltaire’s Candide (1759), Candide leaves El Dorado with a hundred pack-sheep, fifty of them laden with gold and jewels, materials which are not prized in El Dorado. It might be considered an extremely easy heist. Unless it is an extremely difficult one: did Candide fail the heist by deciding to leave this blessed place, partly motivated by love of arbitrage? Was Candide a victim of a kind of switcheroo?

Many classic utopias have a quasi-heist-like structure, in that utopia is very difficult to get into, and contains something very precious (a set of marvellous institutions and norms) which the visitor takes away with them … or do they? In a twist, the priceless haul of Bellamy’s Looking Backward: 2000-1887 (1888), or of Charlotte Perkins Gilman’s Herland (1915), may evaporate when you try to unpack it back at the den. There is plenty of narrative potential in the notion that the same thing might have a very different significance within a utopia or outside of utopia (see Doug Geisler’s interview in this volume).

A crew of daring scoundrels might attempt to steal the source or secret of a utopian society’s flourishing and happiness. But norms and institutions are nonrivalrous, that is, taking them from utopia does not mean depriving the utopians of them — right? Perhaps this is not always the case; the functioning of norms and institutions may sometimes depend, in subtle ways, on their material scope, on their borders, their neighbours, their rivals, their others. Or imagine if the task of the players were to ‘steal’ their entire worldview and mode of societal organization, requiring the group to fully immerse themselves in another way of life to understand and to embody its functioning … and yet somehow not stay there forever.

There is also the theme of contamination of utopia, e.g. as explored in N.K. Jemisin’s ‘The Ones Who Stay and Fight’ (Lightspeed, 2020). There is perhaps resonance here with the reverse-heist or inception: the visitor who brings a catalyst of catastrophe, usually in an anti-utopian fable about the supposed frailty and perhaps futility of building a society substantially better than whatever the author is using as a baseline.

The emphasis on stealth in Doug Geisler’s union organizing game Beat the Boss (2020) gives it a somewhat heist-like aspect. More broadly, TTRPGs offer a space for creating dialogue between heists, crawls and a range of politically-inflected models of taking what you want: enclosure, expropriation, seizure, occupations, squats, appropriation, liberation, recuperation, arrest and de-arrest, incarceration and decarceration, wage theft, rights erosion, extraction of surplus value, mass trespass, or leaving the country you live to go set up a new country. What would it mean to crawl the means of production? To claim the right to roam within the framework of a heist narrative?

After Jean-Joseph Rabearivelo

ay you

stood there nude

you’re mud 

& you won’t forget

only really

you’re a foetus

of labouring dark,

on lunar milk

binging 

by this low wall

inch by inch

into full bole life

crossed over

by flowers’ dreams

& summer sleeps’

perfumes, just

to feel, believe that

from your feet

push out, race & 

worm roots,

& snakelike 

seek your deep 

& wet 

source to slake, &

& already 

bind you to it, you,

O tree alive unknown

untaxonomized

who forms the fruit

you will yourself forage

the bone of your crown

within your hair

that the wind plays with

hides a nest

of immaterial birds

and when you come to bed

and I acknowledge you

my errant brother

your touch, your breath

will awaken the flutter

of mysterious wings

right up to the edge of death

 

Why isn't science fiction interested in AI?

Since the century began, there has been a remarkable surge in AI research and application. This has mostly involved AI of a particular kind: Machine Learning (ML), especially Deep Learning. In brief, ML tends to place much less emphasis on carefully curated knowledge bases and hand-crafted rules of inference. Instead, ML usually uses a kind of automated trial-and-error approach, based on a little statistics, a lot of data, and a lot of computing power. When we hear of AI transforming journalism, healthcare, policing, defence, finance, agriculture, law, conservation, energy, development, disaster preparedness, supply chain logistics, software development, and many other domains, the AI in question is typically some form of ML. 

Despite the click-bait title to this post, AI is extremely prevalent theme of recent science fiction. Isn't it? Well, that depends which AI. Science fiction has been curiously slow, even reluctant, to reflect the ML renaissance. Until quite recently, ML research has tended to de-emphasise anthropomorphic Artificial General Intelligence. Instead it has emphasised domain-specific AI applications. Examples include Snapchat’s AR filters, Google Translate, Amazon Alexa, Tesla Autopilot, ChatGPT, MidJourney, platformized markets like Uber and Airbnb, the recommendation engines that drive Netflix and YouTube, and the curation of social media feeds. 

As a comparison, in May 2023, Science Fiction Encyclopaedia entry for AI still tellingly states: “Most writers would agree that for a computer or other machine of some sort to qualify as an AI it must be self-aware.” Over the past decade, science fiction about AI has continued to coalesce around questions such as: Is it possible for a machine to be sentient, to experience emotions, or to exercise free will? Between humans and machines, can there be sex, love, and romance? Will our own creations rise up against us, perhaps by departing from the rules we set them, perhaps by applying them all too literally? Could an AI grow beyond our powers of comprehension, and become god-like? And what might the oppression of sentient AIs teach us about colonialism, racism, misogyny, ableism, queerphobia, and the systemic treatment of some lives as morally more valuable than others? 

Whether or not these questions make for good stories, or are interesting questions in their own right, they are not tightly integrated into the realities of AI research. This disconnect between science fictional AI and real AI is also reflected in science fiction scholarship. AI Narratives: A History of Imaginative Thinking about Intelligent Machines (2020) is a recent collection of critical essays on AI and literature. While frequently compelling and insightful within its chosen scope, it barely mentions Machine Learning. Terms such as bias, black box, explainability, alignment, label, classifier, parameter, loss function, architecture, or supervised vs. unsupervised learning, appear seldom or never. (I think there are two, maybe two-and-a-half chapters that are clear exceptions). 

Of course, there are some stories that engage deeply with Machine Learning as it is actually practiced. My impression is that these stories remain rare overall, and that they have yet to coalesce into their own richly intertextual conversation about Machine Learning. Some promising counterexamples emphasise 'the algorithm' or 'the platform,' rather than AI as such. They find some storytelling space where a new discourse intersects with an old one: where Critical Data Studies meets the old science fictional delight in robots rigorously following rules, and the humans that might get ground up in those unstoppable cogs. However, even in their more critical moments, many such stories are prone to reinforce the political and ethical framings preferred by tech companies. We can speculate why this might be the case. The economic conditions of their production are worth noting — is there a preponderance of storytelling funded by think tanks, academia, tech companies and tech media, perhaps? Or perhaps there is a sort of discursive predisposition at play, related to the amount of energy it takes to speak outside of the established science fiction tropes. Having laboriously disentangled themselves from questions like, “Please may I have an AI girlfriend?” and “Crikey will I get an AI God?”, are these stories too exhausted to escape from questions like, “How can we balance the need for training data at scale with the privacy rights of individuals?” and “How will the widespread adoption of AI and automation impact jobs and the economy”? Such questions may need to be posed in some contexts, certainly. But they also carry deep techno-solutionist and techno-determinist assumptions. Science fiction could do better!

Writing in mid-2023, there are signs that some aspects of this situation may soon shift. A more recent critical collection, Imagining AI: How the World Sees Intelligent Machines (2023), which does solid and timely work in challenging Eurocentrism in literary and cultural AI, does pay a little more attention to Machine Learning. Even if writers have been ignoring Machine Learning, Machine Learning has not been ignoring writers. And now OpenAI’s ChatGPT is creating an unprecedented level of conversation in online writing communities around Machine Learning. Very recently, Science Fiction Writers of America collated on its website over fifty articles and posts written by its members on the topic of using AI in creative work. Prominent science fiction magazine Clarkesworld recently closed to submissions after getting inundated with ChatGPT-generated stories. The window for limiting global heating to 1.5 degrees, agreed in the 2015  Paris Agreement, is more-or-less closing now, and questions are being asked about the carbon cost of computationally intensive Machine Learning (Vicuna is being touted as a lightweight ChatGPT alternative). Hollywood writers are on strike about, among other things, AI. And in the midst of a messy public rivalry between Google and Microsoft, we are witnessing a sort of convergence of discourse about (the social implications of) Machine Learning with older sci-fi tropes: AGI, Singularity, superintelligence, x-risk. 

Whether or not we are at a turning point, it is certainly a moment to take stock of the last decade of science fiction about AI and ask: Is it possible that the few narratives that engage fruitfully with Machine Learning do so despite, rather than because of, the distinctive affordances of the genre? Compared with most other discourses, has science fiction been good at thinking about Machine Learning, okay at it, or maybe especially bad at it?

Sean Bonney, Baudelaire in English

 Can't remember when this was from, 2007? Sean Bonney reading Baudelaire poems.

Utopian Crunch

From a glossary-in-progress.

Crunch has at least two meanings in gaming. In game development, it refers to an intense, focused, and often exploitative work period that occurs near the end of a project. Workers are racing against the clock, working long hours with great intensity. This could be the time when everything comes together and final adjustments are applied. It might also be the time when all pretence of polish is forsaken. Workarounds are rushed through, risky compromises are gambled on. The product ships with glitches and bugs still twitching, half-squashed, under quick fixes. To the extent that crunch time is also a period of stress and exhaustion, it may also be a time of unusually unreliable judgment: workers may hallucinate that they are applying the final polish, when actually they are inadequately papering over the cracks.

Game development crunch is, in this sense, emblematic of capitalist accumulation: linked to the capitalist drive for profit maximization, the relentless pursuit of narrowly defined economic efficiency which comes at the expense of workers’ physical and mental wellbeing. As John Vanderhoef and Michael Curtin write:

[...] labor exploitation has if anything increased since the early 2000s, as game workers continue to toil under difficult conditions systematically orchestrated by major publishers that have agilely expanded their production networks and political connections around the globe. Although in some places conditions have improved and in others new opportunities have arisen, most shops are governed by wages, practices, and prejudices that undermine common assumptions about game development as an elite creative or IT career. Moreover, the possibility of reversing this overall trend is profoundly uncertain, given the challenges of building a reform movement in an industry where most workers are isolated in cubicles, anxious about job security, skeptical about organized labor, and susceptible to illusions that the indie sector might offer their best hope for deliverance.

Yet this crunch also evokes the logic of the exception. In game development, crunch time represents a state of emergency, where workers must put aside their regular lives and dedicate themselves to the project’s completion. It may have been planned for months, but it is still treated as a crisis. Giorgio Agamben’s 2005 State of Exception argues that the state of exception is a critical aspect of modern politics, enabling governments to suspend everyday laws and deny legal rights during times of crisis. However, the state of exception for Agamben is not really a discrete event that ‘interrupts’ everyday life, but a pervasive condition that governs modern society. Agamben argues that to find the paradigmatic example of modern power over life, you need to look at the camp.

[T]he state of exception separates the norm from its application in order to make its application possible. It introduces a zone of anomie into the law in order to make the effective regulation [normazione] of the real possible [...] the norm [is] able to refer to the normal situation through the suspension of its application in the state of exception.

Achille Mbembe also develops the concept of the state of exception in Necropolitics (2019 [2016]) and other works. For Mbembe, as for Agamben, the state of exception is not really an exception, but rather an enduring and widespread dynamic of contemporary governance. He offers the plantation and the colony as quintessential manifestations of the state of exception, a form of power which goes beyond Foucault's droit de glaive and biopower, and which racialises and instrumentalises human life to the point that those that suffer it are poised between life and death: the walking dead.

So what about game development crunch? Is it an exception to the way things are normally done, or an expression of the way they are normally done here? There are some obvious themes we might touch on here, such as games companies who have a continuous 'crunch culture'. There are also perhaps less obvious ones, such as the ideological role of military-themed games; or the rivalry for control of minerals and other resources for device manufacture. These might illuminate crunch time as an exception that is not really an exception, but rather the very visible part of something semi-invisibly pervasive.

While acknowledging these connections, we might also consider whether crunch times may gesture towards genuine alternatives to capitalist extraction, insofar as they may be times when the standard extractive apparatus of capitalism (money and bureaucratic authority) cannot meet capital's needs, and the more-than-capitalist world must be appealed to. That is, you cannot just pay someone to do that to themselves, so other motives for making things come into play, in however a distorted and corrupted form.

Let's return to that idea in a moment. In the context of tabletop roleplaying games, crunch also has another meaning. It describes a design style focused on complex and detailed mechanics. A crunchy tabletop roleplaying game may involve a lot of arithmetic. It may also involve large menus of options with fine-grained mechanical differentiation, and therefore a lot to memorise or to look up. It may also suggest the existence of many subsystems to handle different sorts of narrative situations (see Subsystem). Examples of games with a reputation for crunch include Shadowrun, Rolemaster, and Phoenix Command.

This second sense of crunch may contain some implied claim to realism, comprehensiveness, or abundant variety. It is also a kind of aesthetic category. Crunch offers an appealing depth to players who love to tinker with intricate systems. In this sense, the term has faintly positive associations. True, a player might say, “I don't like crunchy games.” However, generally crunch is a way of describing something that, if you didn’t like it, you probably wouldn’t call crunch. Instead you would probably describe the rules as too long or complicated — or you wouldn’t describe them at all, because you decided not to invest time in learning them, and so wouldn’t have any particular views about them.

While games might be ranked as more or less crunchy, there is also an understanding that the experience of crunch is something that might diminish with familiarity: to someone who has only ever played John Harper’s Lasers & Feelings, which features two stats and one resolution mechanic for every situation, a game like D&D 5e or Lancer might at first feel very crunchy. The crunchiness of some games, perhaps, is more resilient than the crunchiness of others: the experience of crunchiness does not always diminish, or at least not at the same rate.

For the utopian, crunch might arouse curiosity because it does not look fun, yet it is fun. This speaks to the utopian critiques of scarcity thinking, and the utopian concern with hidden plenitudes, with the possibility of vast pleasures tucked invisibly inside paltry resources, just awaiting the right reconfiguration. Things that don't look edible, but are delicious. Sources of wellbeing and wonder that are less than obvious, because they are being hogged by tiny elites. In a similar vein, activities associated with crunch — poring over customisation options, carrying out calculations, meticulous bookkeeping — have a reputation for being boring, a reputation which is not entirely unfair, but also not the full story. Utopian writing and even utopian society itself has likewise been denigrated as dull (see Boring). 

More subtly, perhaps there is some affinity between crunch aficionados and leftist culture warriors who hold that developing a systemic understanding of something need not necessarily spoil engagement with and enjoyment of that something: that it is possible to be at once immersed in a world and immersed in the rules that produce that world (and to think critically about how each of these emerge from and relate to the real world). This might be described as a ‘culture of systemic analysis,’ often conspicuous, for example, in contemporary culture war clashes around the right’s cherished belief in the possibility of apolitical games, narratives, and art.

What then is the relationship between crunch and immersion? It seems it is not straightforward. Crunchy games are sometimes contrasted with rules-lite or narrative-heavy games, which prioritise storytelling and player agency over complex mechanics. Perhaps players who gravitate towards crunchy games enjoy the strategic challenges, the problem-solving opportunities, and the chance to explore combinatorial wildernesses for local optima and other emergent phenomena. And perhaps players who are averse to crunchy games feel that clearing away a dense fog of numbers frees them to shape story and character. And both groups may well invoke the term immersion to explain their relationship with crunchiness. Are they describing different kinds of immersion? Are the crunch devotees immersed in the narrative and the system, whereas the crunch skeptics are immersed in the narrative alone? Some crunch devotees may argue that crunchiness does not displace narrative, but supports it (or perhaps supports particular kinds of narrative experience). Others may contest that there is any necessary or strong connection between crunch and narrative: instead, we might imagine a quadrant matrix, with an x-axis from crunchy to rules lite, and a y-axis from narrative to non-narrative (some might say, traditional or OSR).

Such questions could be interesting in themselves, and they also suggest another interesting line of enquiry: is crunch something that transcends games? Is utopia crunchy? Charles Fourier’s utopian designs, for instance, might be described as ‘crunchier’ than Edward Bellamy’s, and Edward Bellamy’s as crunchier than William Morris’s. Ursula K. Le Guin’s The Dispossessed might be considered a meditation not only on creeping centralisation, but also on creeping crunchiness (proliferating committees with proliferating conventions), within a syndicate anarchist society that is ostensibly without laws.

How crunchy, for that matter, are the societies we live in today? Social theorists including Niklas Luhmann have in effect explored this question. In Luhmann's framework, modernity has seen subsystems within society becoming increasingly distinct and specialised. This division of labour allows each subsystem to develop its own unique logic, rules, and communication codes. These subsystems can include areas such as politics, law, economy, education, and religion, among others. Subsystems can institutionalise symbolically generalised communication media, making it more difficult for uncomfortable communications to be rejected -- money is perhaps the clearest example (money tells us "do your job" in a way that is harder to refuse than mere promises). In modernity, according to Luhmann, society has transitioned from a simpler, more homogeneous structure to a highly complex, differentiated one. This shift has led to the emergence of a multitude of interdependent but autonomous subsystems. Each subsystem operates with its own internal logic, and is functionally differentiated from other subsystems, meaning they each serve a specific purpose and contribute to the overall functioning of society. Luhmann saw this differentiation as both a strength and a challenge. On one hand, it allows modern society to address complex problems with specialised expertise, enhancing overall efficiency and adaptability. On the other hand, it can lead to difficulties in communication and coordination between subsystems, as they may have different goals, values, and perspectives. There is potential for fruitful dialogue between the theory and practice of tabletop game design and social theory about differentiation and specialisation. Such dialogues might even explore the paradoxical notion of law-without-state, of utopian law.

*

A game designer who is adding crunch may occasionally feel, paradoxically, that their game grows less crunchy: the more one attempts to model the universe in detail, the more the game feels like a model, filled with abstractions and simplifications.

A game with little crunch, we might say, will have a mechanic to determine who wins a fight; a game with a lot of crunch will have a mechanic that allows you to target head, heart, hand etc. separately, to factor in the weapon used and the level of proficiency, the proximity to the target, the wind speed, the sun in your eyes or in theirs as they duck or dodge, and so on. Again, it is all quite relative: to a table of pianists, it may be absurd that a hand is not differentiated into at least five distinct targets. To a table of hand surgeons, all familiar with a range of hand traumas and their treatments and prognoses, it might make sense to separate damage rolls to the flexor digitorum produndus and the flexor digitorum superficialis.

Jorge Luis Borges imagined a Celestial Emporium of Benevolent Knowledge in which animals are divided into those belonging to the Emperor, embalmed ones, trained ones, suckling pigs, mermaids, innumerable ones, etc., high level classifications which feel perfectly natural and self-evident to the authors of the taxonomy. What might a TTRPG manifesting such Borgeian crunch be like? Where should one subsystem end and the next begin? What ought to be considered a special case of what else?

*

What about crunchiness as an aesthetic experience: might this too be experienced outside of games? For example, might the office worker who is entranced in spreadsheet construction be experiencing something crunchy? And if so, could there be ways in which these two senses of crunch — (a) game developers burning the candle at both ends, (b) players losing themselves in intricate rule-sets — are connected? 

Crunch, crunch: certainly, each sense of crunch seems to embody a notion of intense commitment to the task at hand and a desire for immersion within it. Immersion in each case is also figured as something quite intricate and textured. In other words, we are not really talking about immersion in some oceanic realm of primordial unity, where boundaries are fluid and ever-vanishing. We are probably not talking about what the poet Lisa Robinson describes as ‘a luxuriously distributed lubricant, an enticingly shimmering and moving fabric, a shared yet contested décor.’ Rather, the realm in which crunch immerses you is filled with precise relational details, details that are at once trivial and all-consuming — concern with correctness of syntax, conformity with procedure, with how this line of code relates to the next, with how the scattered dice add up or subtract. If there is a loss of self here, perhaps it is more Apollonian in nature than Dionysian; drawing on Friedrich Nietzsche’s dichotomy of Greek tragedy:

Whatever rises to the surface in the dialogue of the Apollonian part of Greek tragedy, appears simple, transparent, beautiful. In this sense the dialogue is a copy of the Hellene, whose nature reveals itself in the dance, because in the dance the greatest energy is merely potential, but betrays itself nevertheless in flexible and vivacious movements. The language of the Sophoclean heroes, for instance, surprises us by its Apollonian precision and clearness, so that we at once imagine we see into the innermost recesses of their being, and marvel not a little that the way to these recesses is so short. 

The notion of jouissance, associated with Freud, Lacan, Barthes and others, may also help to illuminate crunchy immersion. Roland Barthes’s distinction between plaisir and jouissance turns on conformity or nonconformity with norms: for Barthes, plaisir or pleasure arises from engagement within established codes and reinforces established subjectivity, whereas jouissance or bliss comes from exploding those codes, and tends to disrupt and transform subjectivity. Crunch might then suggest a third option, or at least confirms the interconnection of plaisir and jouissance: it is scrupulous obedience to the constraints of coding syntax that opens up vast vistas of representational possibilities. Jacques Lacan’s jouissance defies any brief summary, but there are certainly resonant themes to do with ‘forbidden’ or inexpressible modes of enjoyment, and pain-as-pleasure, pleasure-as-pain. Is there an erotics of submission to game mechanics, with an immersion or flow-state comparable to entry to BDSM subspace? It may be a fruitful analogy, insofar as bottoming is more a transformation of agency than it is a relinquishing of agency.

Each kind of crunch is also associated with time behaving strangely. For example, during game development crunch, there is not enough time to do everything that must be done, and yet it must be done (and sometimes even can be done). Likewise, crunchy mechanics in TTRPGs may slow down gameplay as players consult rulebooks and perform calculations. A crunchy game might therefore be one that has more “time per time,” one in which a few moments of narrative stretch out into minutes or hours of gameplay due to all the extradiegetic intricacies involved. Alternatively, we might not think of this ratio as time to time, as such, but rather story to mechanics, as though these were quantifiable: as though there could be a lot of mechanics per unit of story, producing a sense of crunchiness. Whether temporal or something else, such ratios might interest the utopian: the little hidden in a lot, or the lot devoted to the care and elaboration of the little. 

In a 2021 article, Amanda C. Cote and Brandon C. Harris explore how a discursive distinction between ‘good’ crunch and ‘bad’ crunch has been used to undermine the demands of game industry workers and to perpetuate exploitative working practices. Cote and Harris reject the reality of this distinction. They argue for an urgent need for greater unionisation within the games industry, while also cautioning that even well-organised workforces with robust employment rights may remain vulnerable to exploitation via so-called ‘good’ crunch. Analysing articles and talks from Game Developer magazine and the Game Developers Conference, Cote and Harris point out: ‘Numerous articles and talks positioned self-imposed crunch emerging out of developers’ passion as a good thing, giving specific examples of times when developers’ voluntary overtime improved the resulting game.’ 

The construction of ‘good’ crunch is a weaponised utopianism, which can draw both on neoliberal individualism (personal passion, hustle culture, being driven and ambitious, having what it takes) as well as a more collectivist ethos (the company figured as a family or as a community of mutual care, the worker encouraged to work hard and solve problems on behalf of everybody, workers and gamers as part of a wider community with a passion for making and playing games). Cote and Harris invoke Lauren Berlant’s notion of cruel optimism:

The idea that employee passion drives them to engage in ‘good’ crunch acts as a form of cruel optimism, sustaining crunch practices even in situations where there are organized efforts to improve working conditions. Fully reimagining games’ labor systems will likely require developers both to engage in tactics such as collective organizing and to forgo problematic understandings about crunch.

Cote and Harris’s analysis might be further strengthened by acknowledging the grain of truth within ‘good’ crunch discourse. Like much discourse used to legitimate oppressive practices, ‘good’ crunch discourse reflects, in a distorted form, lived realities. An intense period of work to get a game just right, without financial or professional motives, is something that many hobbyist game-makers experience. It is an experience that they share with, for example, a poet tinkering to perfect a poem or a musician with a song. It is a kind of concluding inspiration, the counterpart to the early stage inspiration in which a creative work begins to coalesce within an apparent void. This is not to say that these periods of delighted productive mania are ‘true good crunch,’ entirely protected from capitalist imperatives — leisure-time subjectivity is still shaped by living under capitalism. But when we do experience them, we are not wrong to wonder: Why shouldn’t most or all work be like this? In other words, these experiences point toward the more free, pleasant, interesting, and less alienated labour explored by postwork theory. They might also be suggestive of the many different ways of working that have existed and still exist in the more-than-capitalist world. The anthropologist James C. Scott, for example, contrasts the tempo of life of hunter-gatherers, “punctuated by bursts of activity over short periods of time,” with that of agriculturalists:

These meticulous, demanding, interlocked, and mandatory annual and daily routines, I would argue, belong at the center of any comprehensive account of the 'civilizing process.' They strap agriculturalists to a minutely choreographed routine of dance steps; they shape their physical bodies, they share the architecture and layout of the domus; they insist, as it were, on a certain pattern of cooperation and coordination. In that sense, to pursue the metaphor, they are the background musical beat of the domus. Once Homo sapiens took that fateful step into agriculture, our species entered an austere monastery whose taskmaster consists mostly of the demanding genetic clockwork of a few plants and, in Mesopotamia particularly, wheat or barley.

As crunch is an aesthetic category, it's perhaps finally worth considering yet another sense of crunch: crunch as an auditory and haptic phenomenon. The sudden breaking or fracturing of a hard, brittle material makes a resonant, sharp-ish sound, a crunch. Crunching is similar to, but not quite the same as, crumpling, snapping, crushing, squelching. Nuts are crunched. Bones crunch, sickeningly. Popcorn is too soft to crunch, apart from the unpopped kernels or kernel fragments. Something which crunches has not collapsed with enough force to fly apart, as something that shatters often does. It may well still be in one piece, connected by fragile new hinges. In fact, gravel crunches underfoot, and here the implication is not something brittle bursting, but rather hard material being tilted, rearranged, and ground together. Pine needles, fresh crisp snow. Perhaps crunch is actually ambiguous between a catastrophic collapse and a survivable deformation.

The aesthetics of the auditory and haptic crunch may also carry utopian charge: concerned as utopians are with pressure from outside, and how an enclosed system might respond to that pressure. When utopias, utopianisms, and utopian sparks and seeds are forced (however temporarily) into narrow margins and small nooks, what survives and what does not? ‘Crunch’ in TTRPG contexts probably derives from ‘number crunching.’ Numbers might be thought of as some of the least crunchable things ever: immutable and abstract, they resist deformation and fracture; you can divide one integer by another, and then combine the result with yet another integer, and nothing really ‘crunches’ — or does it?