The Modern Luddite
Why the Creative Industry Can Embrace the AI Miracle

Author’s Note: This series was written in 2023, when gen AI had only recently invaded the public consciousness. In the months prior, Chat GPT had introduced itself to creative workers with all the politeness of an Eldritch horror, hurtling toward earth from an unknown dimension and making liberal display of its most exotic and offensive appendages. The mood in the trenches of freelance wasn’t as apocalyptic as it might have been – most reactions to the new tech were either dismissive, hopeful, or downright excited. But everyone had their apprehensions: What was this AI stuff really, and how would it affect the creative industry? What reason did we have to trust that it wouldn’t replace us tomorrow, next month, or in a decade?

The Modern Luddite was my attempt to understand how the reality of LLM technology came to bear on the nature of human creativity as both personal expression and economic product. And ultimately to decide whether there was any viable way gen AI could be assimilated into the creative industry to the benefit of its working class. A tall order, but I was convinced it could happen.

Since then, the situation has developed. It’s hard to argue that (so far) gen AI has had a net positive impact on the common person, or the common creator – at best, it’s been a mixed bag. Unsurprisingly, the most egregious ethical breaches have been perpetrated by profit-first corporations, some of which have elected to steal the work of artists and writers in order to train their language models, thereby encouraging support of the models rather than the humans from whom they siphoned their value. Those companies should be held accountable for their actions, and the creative community has been justifiably fighting to do so.

But while the bad behavior of vampiric businesses is always a total bummer for those of us with our hearts still beating, it’s the nature of those businesses rather than the nature of the technology they wield which is to blame. That might not be true of any tech in any circumstance but in this case, I think we can separate the tool from its fabricator: AI hasn’t been anything close to a miracle for creative workers, but that doesn’t mean it can’t become one. Clearly, it won’t be a miracle that’s granted to us. It’s one we’ll have to fight for, to create ourselves.

So despite the systemic failure so far to realize the true potential of gen AI, I still believe we can achieve The Modern Luddite’s vision for an industrial assimilation of LLMs which expands the scope of creative opportunity at every level. I hope its investigation of the differences between human creativity and AI generation represents a small step toward executing that vision.

Sorry for the gazelle/segway thing, in the first paragraph. The machines made me do it.
-Alex

Part I – What’s the Problem, Grandma?

So the robots are here. Old news by now, maybe. Programs like ChatGPT and Midjourney have already taken enough laps around the news cycle to make your average track-athlete, proudly flesh-and-bone and lungs-and-muscle, contemplate the virtues of a segway. The standard fear-mongering pieces and opposing optimistic ones have circulated and recirculated like so many electrical currents through the bundled copper veins of a tireless cybernetic gazelle, who’d make even the heatstroke hallucination of a custom-built-for-tuckered-track-runners-rocket-segway look sluggish.

But excuse my overenthusiasm in the metaphors. Too much that last one, I’ll admit it.

Maybe like many creative writers and artists, I’m just anxious about being thoroughly outperformed by GPT 5 or 6, however long from now, and eager – not desperate, meaning… – to demonstrate my human ingenuity. Think you can write like I can? Obsolete, am I? Then let’s see you string something like this together, ya boring clod of circuits… Forgive me the insecurity – again, I’m only human.

By now, most creative types not too caught up in their “personal projects” (or “monthly odysseys to get the rent paid” if you’re like most freelancers) have found enough spare time to investigate the headlines and arrived at an opinion about ongoing developments in generative AI. And to their credit, I think, most of the views I’ve heard have been optimistic, even when guarded. Cautionary tales have spread about math-teachers’ hysterics at the birth of the calculator and the indignance of portrait painters confronted by the camera – not many of the enlightened thereby are willing to join the lineage of Luddites who would have had us scrawling long-division on the dinner bill and waiting months on end for a new profile painting. Progress is inevitable, yeah? Keep up or eat dust. Grandma and King Lud can enjoy their knitting, that won’t stop me getting my socks 3D-printed.

But some kind of baseline anxiety seems to persist anyway, if my earlier inclinations to try and outdo the computer represent a popular instinct. And I don’t think I’m alone on this one – even in the most enthusiastic discussions of the AI revolution, a subtle tenor of spookiness and uncertainty yet remains obvious among louder tones of hopeful awe and cheery futurism.

Because despite all the hubbub around the potential of AI to help us generate ideas – which is certainly present – and revolutionize creative work – which it can, absolutely – it’s difficult to understand all, or maybe any, of the nuances of what these fancy brain-mimics are doing: The computer got smarter again and it’s spitting out content in an instant that, by all that’s good and natural, should have taken much longer to convince from the stubborn inertia of artistic aether – the blank page, the empty canvas – and probably not been “computable” in the first place. Both experience and our deepest existential sensibilities tell us that this ex-nihilo business should require more effort. A spark of divinity, of human instinct or inspiration, perhaps a sprinkle of artistic angst, at minimum an earnest bit of ritual coaxing. No need to be shy – we all have our cute pagan ceremonies for summoning up the next word, brushstroke, melody… Pre-sunrise coffee, stretch, stare out the window; dip your pencils in holy water, smudge your desk with sage and sweetgrass, or what have you…

No such pious practice for the computer, whose function, apparently simple at the highest level, bypasses the care and struggle we presume as animating forces behind all worthy craftsmanship – inside the GPU it’s a crude operation: eat prompt, excrete content.

Sure, for now the AI stuff isn’t as good as the halfway decent human stuff. Certainly not as good as your stuff or my stuff... But without understanding those nuances of the artificialized Process and where it could be headed, creative people and the creative workforce especially, starving artists and side-gig bloggers and graphic designers alike – e-book tycoons and (why not?) even print-published authors not excepted – are staking their financial security, and often their personal identities, on pure blind faith: the intrepid conviction that what they produce is, and will continue to be, distinct enough from AI generations to retain its value as both sellable product and personal expression.

Getting to the point though, don’t worry – that faith turns out to be justified. There is human value that’s impervious to machine imitation, even in the most extreme sci-fi scenarios. And if we play our cards right, coming paradigm shifts in the creative economy might end up benefitting human artists and creators, even the entire industry.

That confidence comes at no small cost – it’s grounded in a better (if still imperfect) understanding of those elusive nuances of the Creative Process Ex Machina. Asking the right questions of the machine-process, we’ll be able to get a sense of just how it competes with the human one, and whether that’s cause for us to take up arms and remind some computers that, like cursed gemstones bestowed finally into a trash compactor, they can, after all, be junked.

So here it is then – the truth of generative AI, presented as understood by a complete non-specialist with no meaningful background in science, math, or computers, but who’s had enough time on their hands (“lives in a van so doesn’t have rent to pay”) to do some relevant reading: