Decentralized Fiction and the sovereign technology stack

brief thoughts on censorship-resistant publishing

(if you share this diagram, please credit me)

FOR IMMEDIATE RELEASE

Seattle, Washington
December 7th, 2024

AMAZORG DIRECT PUBLISHING is pleased to announce the release of HUGTRON, an advanced trust-and-safety artificial intelligence (TASAI) purpose-built to ensure that AMAZORG products adhere to the highest possible standards of content.

HUGTRON was grown in partnership with leading non-governmental organizations and publishers due to the proliferation of unsafe literature on AMAZORG platform and difficulties that human moderators have had in screening out hateful and divisive content, including Potentially Radicalizing Text-based Material (PRTBM).

HUGTRON’s state-of-the-art algorithms can identify instances of hateful and problematic content in all forms of text, including fiction, analyzing it for “red flags” contained in both overt and implicitly-coded terms. For publishers on AMAZORG DIRECT PUBLISHING, HUGTRON can also provide helpful automated feedback in order to facilitate the removal of text that violate our terms of service agreement. Publishers will be given up to three cycles of submission to have a given work approved by our TASAI.

Existing works of non-fiction and fiction will have a 90-day grace period to modify any works that are in violation of this agreement prior to re-submission for approval. Beyond this 90-day period, any books in violation of this agreement will be removed from distribution.

LLM-based content moderation is, in my opinion, one of the most dangerous and freedom-corrosive technologies that will be developed in the near future. Leading edge AI labs are aiming to get something like 1000x the compute into next generation models within the next 18 months or so, and, aside from the implications for generative text (i.e. the technological execution of the prose artist), we must also consider the likely shape of the Algocracy that is yet to come.

I’m not interested in pushing whatever my own normative view of politics is here (it’s not even particularly well delineated), and for that reason, I try to think about authoritarianism in mostly dispassionate, technology-first terms. And it is becoming increasingly clear to me that the memetic devices that are used to justify modern censorship (hate speech, misinformation, disinformation, etc.) are extremely effective and well suited to the task of increasing state power.

Myself, I think of authoritarianism less as an ideologically-determined state of affairs and more like a “deep pattern,” an attractor force that can pull in all kinds of historical pathways toward the same type of convergent system; i.e., an instance of convergent evolution driven by non-ideological forces.

A central concept in my model of modern technological authoritarianism is the idea of the adjacent-possible as well as the centralizing/decentralizing skew of a given technology. For example, we can imagine that small arms led to the adjacent possible of peasant armies, guerilla warfare, mass politics, revolutionary politics, and so on. We can imagine that the printing press had a decentralizing skew, creating massive levels of political instability by virtue of the liquidity it created in memetic markets for different religious subgroups (i.e. religion memes derived from previously centralized modes of religion). This of course led to all manner of religious conflicts and so on.

I’m not going to scope out an overarching theory of technological authoritarianism except to say that I think that these two underlying features of a given technology, of its adjacent-possible and centralizing/decentralizing skew, are ideology agnostic. What this means is that this technology can be “loaded” by any authoritarian apparatus from left to right. There are leftist forms of authoritarianism, rightist forms, patriarchal forms, and matriarchal forms (“The Longhouse”).

At present, the uppermost strata of the US political apparatus and its broader imperial structures are totally ossified. I don’t think reform is possible until we undergo a cascading series of interlocking crises or until some genuine technology step-function happens like genuine AGI. As such, I am completely checked out of absurd theatrics like “electoral politics” in the managed democracy that we live under. I have zero interest in picking up a banner for or against “wokism in politics” or whatever.

What I am, however, keenly interested in, is literary fiction, and its complete emancipation.

At present, we see something interesting congealing in the alt-lit space, and to the extent that it is interesting, it’s because it derives its energy from outside the centralized systems of publishing that exist in some kind of living-zombie state of culture death. There is something substantive brewing here, no doubt.

Yes, alternative literature is meaningful if a small circle of people enjoy it.

But, in my view, it is more meaningful is a larger circle of people enjoy.

What I mean, simply, is that meaning scales with readership. This, I’m sure, will upset many of you, and I do not claim to believe that this is the sole arbiter of meaning when it comes to literary art. But it is an important dimension of meaning. Jake Seliger, a literary blogger who is currently dying of cancer, once wrote that “books exist to be read.” I fundamentally agree with this definition.

I don’t think alternative literature should aim to remain a niche subject.

I think alternative literature should aim to usurp the current cultural space of literary fiction, which is, for all intensive purposes, an elite pastime (cognitively, spiritually, etc.).

Necessarily, this will require scale, distribution, reach.

This is where the problem of centralized chokepoints come in.

Consider the success story of “Delicious Tacos,” a 100% self-made indie fiction author. Yes: he’s been featured on various alternative podcasts and whatnot, but his success is ultimately a function of his grindset. My concern is that his success is also contingent on centralized chokepoints, namely, it only exists under the mercy of AMAZORG (“he who shall not be named”).

In theory, AMAZORG could easily decide that his work is problematic and erase him as an artist overnight.

As political instability heats up around the 2024 election, as the homeostatic burden of imperial management accelerates through a mushrooming sequence of foreign wars, domestic censorship pressures will increase.

Enter the next generation of LLM’s.

In my view, we may have a relatively small window of time before LLM’s are capable of screening out what centralized powers deem to be “problematic” fiction. What protects alternative literature right now is its obscurity. It’s not worth the money to wipe out dissident artists who rely on AMAZORG because its not worth the cost. Human moderation costs money, but if the costs of moderation go down, obscurity will no longer be protective.

If the technology is validated, I expect AI-based content censorship to rapidly proliferate across every major corporate platform, including payment processors.

Why is this a problem?

Can’t you just buy your own printing press? Run a POD machine in your garage?

These kinds of limited solutions will lead only to stagnation and a marginal artistic and cultural impact.

“Freedom of speech means freedom to transact.”

• Punk6529 on twitter

There is an excellent thread by Punk6529 that explains why financial censorship functionally equates to censorship of speech. The movement of money is oxygen for any enterprise of any kind, including publishing books. Without it, alternative literature as an impactful cultural enterprise becomes impossible.

Credit card companies, payment processors, banks, all of these nominally independent entities are in fact de facto captured by state interests. The only thing limiting their chokepoint efficiency is the cost of moderation, which, if I am right about LLM’s, is rapidly trending to zero. We can expect censoriousness to accelerate at every chokepoint imaginable.

As the censorship-resistant financial system par excellence, we undoubtedly need crypto to solve for this predicament.

(if you share this diagram, please credit me)

In the long run, it will not be enough for us to “build our own AMAZORG” in the minimal sense of the term. Even a Shopify store won’t work if Shopify then becomes the payment processor who is censoring you.

Decentralized alternatives to the existing publishing monolith are not currently sufficient to power an alternative literature ecosystem to acquire the level of reach that we need. They must be scaled to the level that they are viable competitors, such that payment friction and checkout friction are lowered to a level similar to that of AMAZORG. They must be scaled so that alt-lit can grow and usurp our current crop of decrepit wordcels.

This usurping is a necessary and moral event in the historical lineage of the text-based narrative form. It is nothing less than nature calling upon your soul to respawn. Understand that this is not a politically-coded aim. It is an aesthetic aim.

Now, I recognize that what I have prescribed is obviously a tall task. I am no wizard of e-commerce, I don’t know how to do this, I just know that it must be done.

As part of this broader project, we will need a parallel financial system, which is clearly going to be crypto. It is the only viable alternative system of transaction once payment processors really begin to ramp up their censorship of cultural commerce, which I expect to be similarly imminent. We’ve already seen the constraints that a company like PayPal can impose on the production of culture. I think it’s about to get 10x, 100x worse.

I don’t think of this as a battle so much as I think of it as a prison break.

My purpose in this post is not to go into any one player with maximum detail, but merely to explain, conceptually, the absolute necessity of these parallel institutions.

Canonic.xyz is taking on the tall task of tackling both of these chokepoints concurrently, and I applaud their efforts.

I frankly don’t understand Urbit but they are also a key player in this space and are building an entirely independent technology stack in an incredibly ambitious way. I hope to learn more about what they’re doing down the line as I continue writing about decentralized fiction as an ecosystem that flourishes on top of a genuinely sovereign technology stack. Already we see alt-lit magazines like

mars

&

Isaac Simpson

building a readership on Urbit, and this gives me reason for great optimism (even in spite of my own personal reliance on SubStack).

I am excited and energized by what we are going to see, and I will be writing more about it shortly.