AI Slopaganda: A Logical Conclusion of Fascism
To the point: Why is there so much conservative, reactionary, fascist AI-generated content online, and what do we do about it?
The fact that AI content may be entirely slop is of no concern to those who create, promote, and distribute it. Capitalism, if it ever did at all, no longer demonstrates that competition improves products or services; enshittification and rent-seeking are more effective strategies for increasing profits. As with capitalism, so with fascism. It is better for the Fascist to rule over a micro-domain in the information sphere than to educate. Throwing slop at a population that they consider beneath them creates compliance; if the fascist promotes critical thinking through challenging encounters with actual art, they will only undo their propaganda.
An uncritical audience that merely seeks to confirm its own pre-existing biases and hatreds will hardly worry about the quality of the AI-generated gobbledygook country songs celebrating the racist Charlie Kirk, so long as the singer sounds like a “good old boy” and goes through the Litany of American Redneck Paraphernalia (cheap beer, a dog, a country girl, a truck, boots). A viewer who cares only about the superficial characteristics of art, design, and media will certainly not consider, let alone worry themselves over, the overtly white nationalist overtones of "tradwife" TikTok content, despite that content often undermining the propaganda via its production by women who earn salary-level incomes (or more) as influencers.
Why is fascist AI slop entirely unoriginal, beyond the fact that AI itself can only copy and make pastiches of what it has ingested? I must emphasize here a critical point: Fascists fundamentally lack empathy and creativity. Consider the large number of failed-comedians-or-screenwriters-turned-conservative-commentators: Ben Shapiro, Steven Crowder, Jeremy Boreing, the list goes on. Having created nothing but derivative work, they never saw success, and blame their failure on a “woke” cabal silencing them, rather than their own inadequacies.
Given the ease with which "creating"AI-generated art can now produce content that looks real enough to win art shows, it is easy to see how AI can give the fleeting illusion to such philistines that they, too, are artists who can produce worthwhile items of aesthetic value. And if one thing fails, they can simply generate something “new, ad infinitum, until one sticks, while we render our oceans dry and our planet barren.
Conservatism and fascism both also rely on the constant re-packaging of the old for its own purposes in a world that has progressed socially and technologically (e.g. the absolutely grotesque DHS Twitter account). This new world instills in the conservative, and the fascist, a fear of being challenged by the concomitant new ideas. Their propaganda must repeatedly emphasize the sameness and continuity of the ideal future state with an idealized version of the past that never really existed. Unable to promise concrete benefits, they evoke images of the past with sufficient modification for ideological means, and make the false promise of “like that, but better, and with indoor plumbing.”
This reveals itself particularly in the current wave of tasteless re-purposing of American World War II propaganda for the racist aims of the Trump Administration when, obviously, they themselves represent the very same enemy against which that propaganda was designed. World War II was perhaps the US’ most justifiable conflict. But complexity does not matter to the fascist. No matter the underlying reasons the US joined the war (for protection of capital and markets with the more idealistic aims only later added into the collective memory); no matter the mass entry of women into industry, without which armament would have been lacking; no matter the vibrant American socialist movements of the period; no matter the unethical dropping of two atomic bombs on civilians. All the fascist sees in World War II propaganda is a white-dominated past in which the US wrought death and carnage. And with a mere altering of some text, those images, deeply laden with anti-fascism, are now used to deliver the exact opposite message.
We can recall the German joke circulating in the 1930’s, originating from a French Propaganda poster, that the ideal German was “Blonde like Hitler, tall like Goebbels, and slim like Goring.” Fundamental contradictions in the propaganda serve to strengthen, not to weaken, fascism’s reach. The increasing atomization of our online identities reduces contact among groups outside one’s own and reduces one’s ability to recognize these basic contradictions. It also allows seemingly organic fascist movements to envelop an extremely broad, overarching group, composed of all of these tiny microcosms of anger and hatred.
The re-purposing of historical anti-fascist and anti-authoritarian messages has already occurred with substantial success in the authoritarian states of Vladimir Putin, Viktor Orban, and every strongman leader going back to Mussolini. The introduction of AI-generated content has allowed for the effortless re-purposing of anything ready-to-hand into nefarious, soul-draining, hateful propaganda that purposely erases the past, destroys public discourse, and denies us the ability to share truth; a possible and increasingly likely future features the inability to seek out truth at all.
The damage done not only to our society but also our environment surely merits an extreme and decisive response. The ecological impact from water usage to the strain on energy grids all over the country (a topic we will write more about in future) should stand as just a few examples of proof. We could place strict regulation on AI, limit its use, and combine that with regulation of social media to prevent the spread of hate speech and other discursive dangers.
Another option could be reproducing the fascist’s methods online: flooding social media with left-wing AI content, even if it is slop (a tactic currently in use by CA governor Gavin Newsom), with the aim of at least drowning out the fascists ideas, if not converting minds (though unfortunately, fairly unethical given the environmental impact). Though, it is true that producing incomprehensible slop does not require environment-destroying AI models: Python scripts can create thousands of silly, garbled texts per minute, and bots can post them en masse (back to the good old days, if you will). New accounts can constantly be created in order to reduce the volume of “real” content on social media, thereby, hopefully, nullifying the algorithm's emphasis on right-wing content.
Perhaps social media is a commons that deserves to be spoiled in order to save those who use it. The elimination of fact-checking and bot control departments may finally work to the world’s advantage - there is a chance that the proliferation of slop could in turn poison the very datasets being vacuumed up by gullible AI systems. There are many ways to destroy an enemy with far greater computational power than is available to the average user. It is up to us to figure out how, implement these methods, and restore the online space free of fascist, anti-human, AI slopaganda.