Near Infinite Content
Surveying "a radically abundant age in terms of [artistic] creation."
In postulating the emergence of a new profession called a “differentiation designer,” who will inhabit a so-called “taste” role whose “remit combines branding, philosophy, product, risk tolerance and creative execution” in order to cultivate AI media output for the masses, writer Robert Capps in a recent New York Times Magazine article suggested that:
as we enter a radically abundant age in terms of creation, we are certain to see a lot more avenues to creative output that don’t involve the same level of craft.
I appreciate the overall tone of optimism in the piece, as well as the execution in searching far and wide to find new economic roles that we may soon inhabit. But I am not sure I can be so sanguine about “a radically abundant age in terms of creation”—the story wherein AI enables previously expensive and time-consuming art to be generated quickly and cheaply. (Capps actually begins his piece by admitting he found a way of generating a comparatively decent version of the thing he published by way of a clever prompt to ChatGPT 4o (by way of OpenAI’s Deep Research tool), but didn’t actually publish that version because of (i) many hallucinations, (ii) an inherent responsibility not to.1 An admission which, in my book, kind of casts a sinister glean over the whole endeavor.)
While Capps acknowledges that “[t]hese [new avenues] will come with pitfalls, yes,” he says they will also arrive with “advantages.” Which are what, exactly? In an age that promises virtually zero friction in the act of artistic creation, are we so sure we are all going to enjoy the inevitable deluge of content (sorry, “abundance of creation”)?
Let’s call this promised future one of “near infinite content.”
The near infinite content worry is something slightly different than two related-concerns: Ensloppification2 (= enshittification, but with AI), and the Dead Internet Theory. Both of those things deserve attention too, of course. (I’m told that currently four of the ten top YouTube channels, and over 41% of all Facebook posts, are now AI-generated. Our brave new (dead internet) world is here.) But if the near infinite content future arrives, it doesn’t necessary mean that the stuff we will encounter3 is going to be slop or fake. Instead, maybe the vast majority of AI content does end up being human-cultivated at some level—and let’s take on Capps’ optimism and suppose that it’s even likely that a good amount of it will indeed be cultivated by humans with good taste. So that some of it will at least be…not bad, I guess.
Ok so, the bullish view here would be that, with so much stuff being newly possible to produce cheaply, more creative gems are inevitably coming our way. (Corollary of increasing the number of chimps banging on typewriters = more of Shakespeare’s plays get written?)
My problem (laying aside various and sundry concerns about the entire task of AI-generated art in the first place) is: in art, barriers to entry are sometimes laudable; and unfortunately, they seem necessary. Such that even if AI merely augments human-driven art, but we still get the near infinite content future because of it, I think we are headed for trouble.
We can look at the question from both the supply-side and the demand-side.
Supply-side.
In a pre-digital America, you would sometimes hear artistically-inclined people complain about, e.g., how no studio would ever be brave enough to fund their movie idea. It was maybe this culture of creative ennui (overproduction of elites?) that prompted Francis Ford Coppola to remark in ~1991: “one day some little fat girl in Ohio is going to be the new Mozart . . . and make a beautiful film with her father’s little camera-corder, and for once the so-called professionalism about movies will be destroyed forever, you know, and it will become an art form.”
The thing is though, the cost to film the kind of home movies FFC has in mind here fell to effectively zero in the decades after he made these remarks. And yet plump young Ohioan Agnes Varda has yet to break onto the scene. Neither the Super 8 cameras nor YouTube ended up turning movies into an art form. The utter ubiquity of the iPhone camera in the developed world did not cause a surge in young cinematic auteurs (unless you think, um, that influencers and live streamers are making art).
The emergence of a new artist seems to instead be a delicate thing, and to not necessarily arise from putting the tools of the greats into the hands of the people. A delicate thing, necessitating near-miraculous acts of discovery, nurturing, and risk-taking. The “gatekeepers” here in fact performing a necessary and vital role. Those gatekeepers not so much functioning as entities who stop people from pursuing their dreams, but instead entities that—funny enough—enable it for a select few.
And as much as many artists have detested their encounters with industry, surely being asked to direct a commercial, or to photograph a new line of furniture, or to sculpt something that would look perfect in the lobby in a certain bank, helped burgeoning (or even mature) creatives develop their skills and capital and begin to dream about what use those skills could be put to. When industry just turns to AI in the first place, these avenues for practice for people who want to pursue the real thing therefore dry up.
It therefore seems obvious that with near infinite content, you probably actually end up with fewer artists. And certainly fewer good ones. And fewer good artists means less good art (in both senses of the phrase).
Demand-side.
It used to be that the slush pile remained solidly out of sight, and with it, consumers of the arts could focus on good, interesting stuff, barely even realizing (if at all) that their consumption happily sat atop a massive pyramid of mediocre work (the majority of which had no chance of ever reaching them).
That version of the world has surely already been exploded by the Internet, where of course anyone can post anything. But in a version of the Internet where effectively anyone can post everything, how confident are we about our collective appetite to sift through it all?
The world in which everyone “writes” and “directs” a movie by prompting AI to do it is probably just one where people want to watch fewer movies overall. The paradox of choice playing out over the Library of Babel means that few people ever really feel like going inside to pick up a book.
And that’s partially what bothers me about the whole “taste” discourse around AI. (A lot of people are convinced that humans will inhabit this important role of applying “taste” to AI output. The view is that, in the age of AI, taste becomes king.) But why even consume someone else’s stuff in the first place, when all that other person is bringing to the table is their taste which is, at best, slightly different than your own? Why even risk an encounter with someone else’s filter, when you could just go the AI yourself and generate your own thing at a similar level of quality—something that you can tune to fit your preferences better than probably anyone else could design it to for you?
And then, supposing you then have these AI-generated things that fit your taste just so, works of “art” which speak to exactly the areas of your spirit that you always felt you were exactly yearning to be spoken to; are you even, like, satisfied? Do you actually feel compelled to go out and consume more? I mean, no! The “artwork” is just there somewhere in the machine, and you can call it up on demand the way you order meals on Doordash. I bet there ends up being nothing truly compelling about heavily-AI-assisted art in a world where on-demand AI art exists—regardless of how sophisticated it all becomes—with nothing really prompting you (as it were) to go out and watch more, read more, feel more. Maybe you yearn for the pure human stuff, maybe you end up accepting a hybrid—I don’t know. But either way it cheapens the experience and thereby lessens the demand.
So, I think the paradox on the demand-side is that, although the digital public indeed demands near infinite content, the existence of a near infinite content world will actually make people less likely to consume works of art.
So I guess I’ve convinced myself that the near infinite content story looks more likely to herald an age which diminishes the truly special rather than a golden age of artistic thriving.4 But Capps’ optimism is at least a little infectious. So in that spirit, I’ll hope that I am as wrong about this one as Francis Ford Coppola was about the portly young auteur in the Midwest.
Paul Gaugin, L’univers est créé (1894).
This Week on AI Twitter
The last thing you hear before the machine takeover. Critical theory to the rescue. Man as AI’s best friend. Project Vend. Critical commentary on the $100m signing bonus thing. Signs of an economic revolution in the making.
His idea that writers are paid only to be responsible for the facts asserted in their work is totally wrong, but that is another talk show…
This would be the dark side of the “not involving the same level of craft” idea.
I guess in this account, both the online and offline worlds become saturated with AI-generated media.
Safe to say that good art, like the human species itself, will never be altogether extinguished by technological onslaught.