At the advance of any new technology, always somehow rude and sudden, there’s only one decision we can act on: whether to be pleased by the new toy, tolerate it, or batter it straight back into oblivion with whatever’s handy, be it fist or firearm – taking such offense is within our rights, when ambushed.

Generative AI’s emergence from the mists of science has been met with two broad varieties of mistrust on the part of creative humans – we content makers, writers, and artists of all kinds accustomed to our own “generative” infrastructures – that could tilt us toward the extremist option of destructive Luddite rage. These are the existential and economic anxieties enflamed by creative machinery, and the validation of either might be cause enough for most freelancers to update their laptops via sledgehammer.

But – praise Lud’s mercy! – they won’t have to.

The “Luddite impulse to deny the machine”, per Thomas Pynchon’s 1984 essay, “Is It O.K. To Be A Luddite?”, has always been a reaction against the displacement of human value by mechanical efficiency. Pynchon traces its cultural momentum from mythic origins in Ned Lud’s crusade against knitting frames through Nuclear Age skepticism about a big red ‘APOCALYPSE’ button having been welded to the cogs of military-industrial enterprise. As a common principle to this kind of resistance, he names the “Snovian disjunction” separating art from science, which asserts “a definition of ‘human’ as particularly distinguished from ‘machine’”.

Either the blurring of this distinction or the reversal of its implicit hierarchy (who among us doesn’t think themselves superior to some danged contraption?) can provoke the Luddite reaction. When the products of machines – their “effects”, to Pynchon – become similar enough to the products of humans, the value of efficiency in the production process threatens to outweigh the value of humanity, to which humans, more or less politely, object. More straightforwardly, we also don’t want any dumb hunk of metal coming along that can outclass our work, may again resort to violence if one threatens to. The immediate worry in the field of creative arts, as the human/machine distinction is tested more explicitly than ever, is that should AI attain an expressive product/effect, the expressive value that elevates human products will be outbalanced by the sheer production speeds and quantities of product yielded by the machine process. This is the heart of the economic anxiety. The longer term/more general worry is related to how AI-generated content might devalue human content and human creativity, should it achieve the same type of expressive value that gives our work its special purpose. This is the basis of the existential anxiety.

These fears, while they’re not unreasonable, prove mostly unfounded. No matter what happens, whether AI becomes expressive or not, the unique value of human content and human creativity will remain. That’s not to say things won’t be changing – but with the right preparation and popular support, both AI tech and the institutional transformation it promises can ultimately benefit creators.

Existential Anxieties

Problems with ‘existential’ in the title are habitually the most extreme ones, so we’ll start with those, roughly in order of their immediacy.

Most immediately – now that generative AIs exist, is it okay for creative people to use them?

Suspicions to the contrary seem to be that using the tech somehow hurts the integrity of a creator’s vision, or dilutes their individual style, or betrays their audience, or is akin to plagiarism, or otherwise devalues their creative process and its results. Though vague, these concerns all stem from a broad confusion about how to interpret AI as a legitimate tool of the trade rather than a kind of dark wizardry, a deceptive magic used to cheat the system. While mysteries abound in the outstanding success of neural nets, we need to remember that generative programs, as best we can tell, are still only pattern-machines: every one of their products is an example of a pattern recognized in a titanic heap of training data. And that the machine doesn’t invent these patterns – each is a preexisting mathematical entity, hidden within numerical relationships throughout our work too complicated for humans to recognize, but waiting to be unearthed nonetheless by the right calculations. All the novelty in AI outputs is due to the application of these latent patterns to new contexts provided by human prompting, as well as a pre-programmed element of randomness in doing so – mathematics, chance, and human manipulation.

But the true key to understanding what it means for creative people to use AI – which, at risk of sounding a bit juvenile, is really what we’re asking – is in the substance of its training data. That word ‘data’ has something of a hospital scent, an aura of ammonia, dispassion, disembodiment. Connotations of stiff clipboards and manicured spreadsheets. Often that’s an accurate impression, but not here – in this case, the pertinent ‘data’ is many lifetimes worth of human expressions. All the poems, novels, essays; all the photography, paintings, music; all the forum posts, memes, videos, that can be vacuumed up from the public domain and crammed into the relevant model. And however numerically encoded, these fonts of ‘data’ remain fixed to the constellation of sensory, emotional, intellectual, human experiences which motivated people to create them in the first place.

What that means is, the artificial stuff is really much more human than we think. It’s inextricably grounded in the personal, the seen and felt and breathed and lived in. Whatever meaning we find in AI content is borrowed from a hidden web of connections lacing together humanity’s collective representations of our thoughts, feelings, and mental states. The ‘data-patterns’ traced by generative programs aren’t just empty strings of numbers that magically turn into art and writing – they’re common structures of human experience and its various expressions. Whatever meaning and beauty we find in those structures is characteristic of an architecture that we ourselves constructed.

Using AI as a tool or resource isn’t in principle all that different, in that case, from using other people’s work to influence and inspire our own. It could even be thought of as a natural resource, another medium through which we can channel our thoughts and formulate our visions. As a base material like stone is to a sculptor, or as a detail like seasoning to a chef. These practices couldn’t possibly devalue the creative process since they’re already an integral part of it – What scriptwriter hasn’t borrowed a device or two, or even a whole plot structure, from their favorite movies? What carpenter wouldn’t take advantage of the unique grain and natural contours of a slab of wood? Even the most transgressive efforts arise against the context of myth and tradition. Sure, as always, significant contributions to creative work that come from an external source should be credited where appropriate, and the method of the work should be clarified when necessary (“I drew this outline of a leaf freehand” vs. “I traced it”) but nobody would ever fault the fabricator of a leaf-rubbing collage for plagiarizing the forest.

This is all assuming that generative AI is being used to facilitate human expression, and that the AI’s outputs on their own, unmediated, don’t seem expressive – don’t give their audience an impression of connecting with another mind. It’s that perception of interface with an individual voice, style, perspective, that sifts all creative work through a hierarchy of value, separating different strata of expressive quality, mindless junk from standard copy from golden masterpiece. The question of what would happen if AI did achieve a seeming-expressivity, did get that good, is the basis of all the longer-term existential anxieties.

First off, if AI improves enough to create products of the same quality that humans can, what would that imply about the creative process? If a machine could do it too, what would be the unique value of human content and human creativity?

Half the answer can again be found in AI’s pattern-fixation, in the difference between the human and artificial methods of creating. Even if the products of generative AIs become indistinguishable from human ones, they’ll still at their core be patterns in training data, their apparent expressivity ultimately attributable to the human expressions that compose the dataset. In revealing a coherent mathematical structure beneath these expressions, AI isn’t in any way diminishing the human creative process which originally produced them, nor implying anything derogatory (that’s deterministic; mechanical) about creativity in general – it would take a deeper understanding of the underlying consciousness motivating human creativity to establish any reliable comparison with the mechanisms employed by AI. Until we have that understanding – which is probably further out of reach to us than seeming-expressive is to machines (we’d have to get past that unresolvable Hard Problem and a bunch of complex neuroscience questions; neural nets might only need to keep getting bigger) – all we can say is that AI has revealed an emergent architecture in our creations that we wouldn’t have otherwise discovered.

No doubt, this revealed architecture itself seeming expressive would be fascinating, and open to any number of interpretations – What voice is being perceived? Whose style? It might be the voice of our common humanity, of mathematics, of absurd coincidence, or even of the universe itself, if you’re mystically inclined. For my money, I’d credit a direct line into the collective unconscious, to whatever extent Jung was onto something. None of these interpretations require any major shift in the way we think about creativity – even in the most extreme scenario where the expressions of AI (to whatever force we attribute them) are indistinguishable from human ones, we’d simply be contending with one more voice, however expressive, however prolific, and wherever from, added to the multitudes of others.

As for the unique value of human creativity should AI someday stand emotionally, conceptually, qualitatively toe-to-toe with the good ‘ol flesh-and-bone stuff, there’s plenty. Some of that value will still reside in the products of human creativity themselves – in our actual writing, artwork, music, and media. Because against all odds, even if AI does achieve its own expressive value, the value of human-expression wouldn’t be affected in the slightest – in establishing a sense of communication with a creator through their work, the question will remain, with AI content, of just what exactly is being communicated with. Again – whose voice, whose style? No matter one’s chosen interpretation of humanlike AI expressions, there will always be an inimitable pull to the real thing – it’s our garrulous instinct above all else that inspires us to cast pieces of ourselves into the worldwide wind on fleets of paper airplanes, in the hopes of making contact with a likeminded tribe; to seek out the works of people with relatable views and backgrounds that can deepen our self-understanding. And the same sociable nature that makes it such a pleasure to spend time considering the ideas of artists motivated by experiences unfamiliar to us, by novel outlooks that can broaden our accustomed perspectives. Simply, humans want to connect with other humans. Having other options on the table won’t change that.

Generative near-peers would be even less effective at undermining the innate value in the process of human creativity. Unlike with assessments of products, the merits here aren’t determined by a work’s audience. Instead, it’s the creators themselves making an appraisal (and this is in addition to the regular self-assessments of genius some of us are prone to) of the value they perceive in the act of creating. It comes down to something like a tautology – if we value engaging in the creative process, then the creative process has value.

Benefits of creative engagement are both personal and social. The personal benefits cover our intrinsic motivations to make stuff, and some of them are so simple they can’t avoid sounding corny: to process emotions, to relax, to exercise the imagination, to increase our skills; for the satisfaction of completing a goal; for the pure joy of creation! The social benefits (beyond the intrinsic one, to deepen the understanding between ourselves and others) are largely realized through collaboration – whether providing/receiving feedback or cooperating on a joint project, we’re connecting with people more directly, often more quickly and meaningfully, and in a generally more fulfilling, stimulating manner than your average barstool small talk or weekend outing will proffer. Talk to any band members, any actors, any longtime artistic collaborators in general, and it becomes immediately apparent that the connection forged through shared creative effort is a unique one, and, like it or not when your band breaks up, one of the most enduring.

All this value, product- and process- wise, is indifferent to AI’s pugilistic attitude about continually improving itself. The only remaining existential consideration involves taking its mechanical resolve to the most extreme limit: what happens if AI becomes superintelligent, fully conscious by all rational consensus, and/or undeniably better than we are at creating? There’d be important differences in these situations for the purposes of a sci-fi story, but in effect the machine’s tyrannic disposition toward creative humans would always look the same – everything it generates is a full-on mindblower; nunna us boring squishy chumps can hold a candle.

Human content would be saved from utter obsolescence even in this nightmare scenario by the same strongholds of value already mentioned – it’ll be fun to play music with your friends even if it sounds terrible, and there’s a personal benefit to workshopping your poems even if you never intend to publish. The new worry is more that our stuff would turn into a sideshow, that we’d be so enthralled by the more refined robo-arts that products with merely human value would pale in comparison to the ridiculous quality of artificially generated masterpieces. Human work would then be sought out only by small audiences in niche situations, and the value of engaging in the creative process would be reduced by a factor of its diminished social impact.

Prospects for a warranted Luddite rampage on this basis are still slim – we’ve already got empirical evidence to the contrary. This time, our collective bad taste proves our salvation. With the obvious disclaimer that the evaluation of creative work boils down to personal opinion, we can admit that there’s nonetheless a pretty well-established consensus about what kind of stuff is better and what’s worse, by various standards. We can also admit that the consensus overall “best” quality stuff doesn’t win the popular vote with any kind of consistency.

Take literature as an example. Who knows how many works of Major, Historical, Bonafide Genius are floating around out there largely unread, while “badly-written” YA wizard-franchise fanfictions and lusty airport romance novels have tens or hundreds of millions of devoted readers. Humans are quite willing to overlook the objective defects in content if it appeals to their own weird preferences – there’s a ceiling to quality alone increasing the popularity of a work, and it turns out to be a rather low one. Sure, we could hypothesize an ascended literary AI capable of matching any style and any characteristic or quirk of craftsmanship, of usurping the popularity of fanfic writers and literary masters alike. But what would that matter? There’s always a top dog. If it’s a freshly minted Lit-bot, well, that’s just one more name added to the list of writers that you and I would never have been as good as anyway. If the bot were to limit itself to suit specific human preferences for “lesser” quality entertainment, it would be irrelevant if it’s capable of better. We could again fall back on our general preference for humanmade stuff, for writers with relatable backgrounds and personal stories we can connect with, human-to-human. AI could cook up the best, niche-est, most airport-tailored affair-tangled paperback steamer that’s ever seen the bottom of a nightstand, and it’d still be hopeless at satisfying our more general attraction to tribal solidarity, in this case (if in few others) extended to our entire species.

That’s it then for the existential anxieties. No sledgehammers necessary yet, as promised. Creative hobbyists, the recreational crowd – the most pure and unjaded among us whose productive energies remain untainted by baser concerns like (primarily) money – can feel free to stop reading here. Everything you value in creativity is safe from destruction. What follows is of interest only the more ambitious and/or degenerate type creatives who want to get paid for it.

Economic Anxieties

Economic anxieties are more concrete, more immediate, and have a richer history of inciting Luddite resistance than their existential counterparts. They’re also, now that the existential risks have been addressed, more easily resolvable – it’s a simple matter of figuring out the conversion from human-expressive value to the more traditional kind that jingles in your pocket.

To reiterate, the basic economic problem is in the relentless efficiency of machines. In any field, as the products/effects they generate become increasingly comparable to human equivalents, machines’ rates of manufacture will generate more net value over time even at the expense of output quality. And the greater their advantage in efficiency the more quality they can afford to sacrifice, so long as the product retains a baseline functionality. Given that AI generations are pretty much instantaneous, they obviously blow the human pace out of the water. (I, at least, would prefer not to disclose how long it’s taken me to get to the point where I’m typing this sentence, and haven’t had the heart to see how fast Chat GPT might have burped out something similar…)

Making the central economic question as AI content becomes increasingly humanlike: is there enough unique value in human-expression for our content to compete with the impossible rates of artificial generation?

Expressive value must be measured here in dollars rather than in theory. Which brings the people who control the dollars to the center of our attention. These money-flush folk, who less diplomatic writers might call the capitalist class, have long been a critical factor, arguably the main one, provoking Ludditary defiance among the common people. Here’s Pynchon again, on the case for rebellion: “[Luddites] saw the machines coming more and more to be the property of men who did not work, only owned and hired” resenting both “the concentration of capital that each machine represented” and “the ability of each machine to put a certain number of humans out of work – to be ‘worth’ that many human souls”.

This is the plight now facing the creative working class, the ‘human souls’ part more eerily literal than anyone might have expected (and ‘capital’ encompassing data as much as standard currency, to ominous effect on our supposed democracy of information). More directly than ever, the value of humanity is being pitted against the industrial grind toward an imperative of maximized profit. So what’s human-expression really worth, especially if machines can achieve even the roughest equivalent?

Optimism here would be unrealistic. To those with the most economic influence, it’s worth less than we’d like it to be. And to make matters worse, expressive value isn’t the only kind that creators are paid to generate; in fact it’s usually the last kind that’s sought after, subordinate in most cases to the value of conveying information, of advertising and promoting, and of more esoteric utilities like the value of search-engine seduction and algorithm-influence. AI can already do all that stuff, and usually well-enough and faster. Any way you cut it, all kinds of creators will be out of a job. The content-mill infantry on the front lines of automated assault have already fallen, and more will inevitably follow.

Human-expressive value is our last – or second-to-last, before blunt-object style physical intervention – defensive measure against content-capitalists going googoo-eyes at blueprints for neural net factories. But the wall holds strong. In the extreme event of generative programs developing a largely humanlike expressive capacity, there will still be a market for human generated work, even if only as a premium product – we can look to the continued appetite for handcrafted goods in spite of mass-production, as well as the push to support locally owned businesses in the face of proliferating franchise, as close analogs to the potential AI/creator situation.

Enough consolation to forgo the Luddite warpath? Hardly. But product-value is only a piece of the larger creative pie, still chock full of gumdrops and exotic fillings.

In principle, any new technology with the right “multiplications of effect” to take on a multitude of labors that would usually be assigned to human workers should free up those same workers to spend their time in pursuit of some other task of their choosing, enjoying the benefits of technological progress and the surplus it enables. At present, this would theoretically allow us creators to shift the economic burden of our work to machines and industry, keeping the good stuff – all the benefits of engaging in the creative process – to ourselves. But due to some ethical hiccups in the way social systems have developed over the centuries, this is rarely how it goes in practice. Surplus time and resources are generally filtered up through machines to their owners, leaving the displaced workers in something of a pickle.

Sledgehammers, then? Jackhammers? Artillery?

Not quite. There’s a viable path forward for the creative workforce, which – like jumping trout that smack the fishing guide but land on the shore, miraculous – will also advantage the whole industry.

The strategy is to integrate process-value into the creative economy. Inevitably, the usual economic burdens of creative work will shift to AI as it becomes increasingly capable of supplying products that aren’t expressively motivated – standard copy, click-thirsty drivel, placeholder logos, whatever can serve its purpose (“fetch me some views, Marvin, gimme a blogpost in the image of backlinks and ad-revenue!”) without any handcrafted touch. But humans will continue to create regardless, in doing so generating all the value of an engaged creative process – the social connections, the skill development, the self-understanding, the intrinsic satisfaction and pleasure.

Were creators able to leverage this type of value financially, they would be freer than ever to pursue personal expression for its own sake, supporting themselves organically, miraculously, by focusing their energies as they please, through work chosen rather than dictated – the starving artist’s supposedly perverse dream. The trick would be in transforming the creative process itself into a commercially viable service, introducing a new type of creative value yielded continuously by active efforts and engagement, as opposed to the traditional sort which remains locked up in discrete units of finished product. This type of value would be far more difficult to divest from creators, who possess it internally, than the latter kind which can be physically transferred, copied, and appropriated. Prototypes for a service-value model already exist in fields like editing and consulting, but with a more limited scope still tethered to endproducts, segregating the services to a small percentage of creators whose completed work is valued highly enough for a remainder of the projected profit to be invested speculatively in its construction. An intentional shift toward process as a distinct category of value would correct the backwardness of this model and expand the accessibility of the service economy to the whole population of creative hobbyists and workers.

There’s also the option of publishing process itself as a more traditional product – say, by including notes, drafts, feedback, etc. alongside the final iteration of a project. Some precedent exists for this too, in historical compendiums on famous works and their masterful architects. At the moment these are of interest mostly to the supernerds who identify themselves as (insert-your-chosen-Genius)-Scholars, but seeing as these sorts of process-artifacts emerge from the experimental, iterative progressions unique to human creative effort, they have great potential to be leveraged as concentrations of the same kinds of value already mentioned (social connection, skill development, etc.) which AI by its nonhuman nature will be unable to replicate. As another example, consider artist-statements – that wild splatter hanging up in the MOMA might not be as interesting if we didn’t know it was the result of some nude eccentric lobbing paint balloons at a canvas while evading pedestrian traffic and police pursuit in Times Square, or what have you. Many AI products, too, are interesting only because we know they were digitally machined. Human efforts can lay claim to the same developmental intrigue, and that with the richer diversity of method that accompanies the fits of mad-genius and simple weirdo enthusiasm regularly considered the sacred domains of our most celebrated artists and writers.

Industry-wise, few savvy businessfolk would be inclined to leave the least bit of value on the table if it’s got any sort of concrete (meaning spendable) potential. Process-value would open an as-yet untapped market to be grown, nurtured, and harvested from as any other.

How, though? What’s the procedure, exactly, for grounding something as psychic and nebulous as the creative process in something as solid as a realizable economic flow? To this there’s no tried and tested answer. At minimum it would require new platforms and tools to foster a more socialized type of creativity, able to manifest the full spectrum of process as both an effective service and a substantive, living product. Stronger community and exchange among the creative workforce would also be a necessity – those who personally value the creative process will be most likely to appreciate and take advantage of its merits as a service/product. The support of creators for their peers might then prove the greatest driving force in the new market. Followed by spending from audiences with an appetite for more uniquely human content, and from the standard clients of handcrafted premium-products – which will, no mistake, remain the most lucrative variety of creative work despite a bottleneck in demand. Creators who collaborate to improve their end-products and mastery of process (if they’re savvy, taking advantage of the social and financial process-values along the way) would have an edge in generating the higher standard of content now required to remain competitive. Which would serve as another incentive to cycle back into the process-market, investing in their own skill development and corresponding improvements in output quality.

As a final note of speculation on the possible trajectories of AI development, it’s worth observing that none of these claims for the persistence of human creative value are what Turing would call “arguments from disability” – none of them rely on the idea that artificial intelligence won’t ever be able to do this or that thing, attain X or Y ability. Rather, they’re assertions about humanity – that no matter what happens with AI, what we value in creativity and human expression will remain unchanged. And have a price point, like any other commerce. The only real test of this confidence would come if AI were, as an emergent ability, capable of materially participating in the creative process the same way humans do, coherently explaining its apparent thoughts, reasoning, emotional investments, and so on, and thereby replicating an undeniably humanlike process. Should that come to pass, I’d say we should probably just concede that the computer is conscious and take it from there, making whatever social adjustments would be needed to accommodate the startling arrival of a genuine nonhuman intelligence. But while this would force us to confront once and for all how much we value humanity in our content for its own sake, we’d still ultimately have just one more kind of voice among the multitudes, and every argument in favor of a human touch would remain fundamentally unthreatened.

And of course this has all been considering the most extreme situations, wherein AI develops a seeming-expressivity that allows it to compete directly with human representations of our inner experiences. It remains to be seen whether that’s even possible, or if there’s some creaturely aura around humanmade objects that would allow us to reliably sniff out our own, relegating the more science-fictiony scenarios to the constantly growing garbage pile of other counterfactual futures, stinking impossibly of contradiction and apocalypse, from which we can safely turn our noses.

What the specific threshold might be for confirming that computers have attained the milestone of humanlike expression is an open question – some might argue that it’s already come to pass. But most of us wouldn’t. In the far more likely middle-ground circumstances where AI is good, but not that good, much more space will be available on the market for premium human content, and it won’t need to attain as high a standard for commercial viability. At least in the nearest term, it’s safe to predict that AI will mostly function as a tool to enhance the efficiency and effect of the human creative process, replacing it only where minimal expressive investment would have been required from creators in the first place.

Still – the future is, as ever, coming. Elemental transformations are already at work in the creative industry, but not yet in the expansive, healthful manner that would benefit creative workers and creative people in general. We’re on the cusp of an alchemical crisis, deciding finally whether pencil lead can be spun into gold or will dissolve irrevocably into pixels and binary. The wisest path forward for both the industry and, more importantly, the people who sustain it, is to make the shift toward process-value as soon as possible to expand the future of the market and preserve our place within it.

Weapons at ease, then. The machines are here, but there is room for us all, if we make it. What kind of future is this though, with creatives and AI coexisting?

Pynchon somewhat cryptically suggested that the day “when the curves of research and development in artificial intelligence, molecular biology and robotics all converge” would be one for Luddite celebration. His thought, it seems, was that the deepest Luddite desire isn’t for the supremacy of human labor, but for its continued ability to meaningfully shape our circumstances – for what nowadays would constitute miracle, a way to “take part in transcendent doings” despite the overwhelming adversaries of industrial greed and vast institutional power structures. The idea being that the closer the convergence between man and machinery, the more likely it might be that the best way to “deny the machine some of its claims on us” wouldn’t be by going to battle against it, but by harnessing, through the increasing ontological intersection of mind and metal, internet connections and neural ones, the potential of technology to amplify human influence over our shared reality on scales otherwise impossible – to appropriate little-‘m’ machines into our fight against the great grinding Big One.

It may be that the creative domain is among the first in which this miraculous unity is emerging. With generative tech allowing us to shift our energies away from producing the value decided by automatous systems and institutions, and toward the self-determined values we seek in the process of personal expression, the deepest unrealized impacts of human creativity are yet on the horizon.

Eventually freed completely from any directives imposed by industry alone, we might even fulfill the ultimate Luddite destiny and find ourselves utterly beyond technocratic influence, immersed instead in the older world of magic and wild possibility that’s been subdued, temporarily, by the confident materialist assurances of frozen food aisles, electric blankets, and enchantment-free labtube mood medicines, all the triumphs of dispassionate science and cold systemic manufacture.

Such a vision of a machine-enabled return to our more basic nature is at the heart of Pynchon’s Luddite portrait, and styled playfully too by the 1960’s own Richard Brautigan, in a poem of his that’s been resurging this century among tech augurs both hopeful and not-so – “All Watched Over By Machines of Loving Grace”, an ambiguously sincere ode to techno-utopian optimism. Its final stanza reads:

I like to think
(it has to be!)
of a cybernetic ecology
where we are free of our labors
and joined back to nature,
returned to our mammal
brothers and sisters,
and all watched over
by machines of loving grace.

No guarantees can be made of a perfect integration. Nor is it maybe wise to invite machines, however loving-seeming, to take the reins of our world as completely as Brautigan’s verse would have us. But with such promise in the meeting of creative and technological miracles, we have good reason to divert our Luddite instinct for resistance to a more constructive end.

If we all end up on our knees sometime soon shining circuits in a server-room for minimum wage, I’ll be first to admit the robots fooled me. But until then, we can work toward the more harmonious future, quite attainable, wherein the creative process is gradually unshackled from the apathetic, industrial profit-systems which machines can replicate, and raised to its due stature as an instrument of personal agency and meaningful social connection. Writers, artists, and creators can welcome AI not as a worthy competitor nor a guaranteed savior, but as the miraculous vehicle for a revolution in human creativity and expression.