In true Millennial spirit, I find myself on occasion staying in for the weekend to wander around cyberspace, absorbed in the glow of a triple monitor setup and using the words of my fellow online companions to keep me company. On rarer occasions still, a stiff drink will accompany us on our escapades and mischief, leading to cherished moments of closeness and unity w/ someone who, by all appearances, is a perfect stranger and yet feels so familiar that you’d swear you had been by each others’ side for at least a decade. Over the course of one such night romance and ex romantic partners became the subject of conversation, leading to an acquaintance of mine saying1:
“But also if that person is no longer in your life you kind of have to decide on what it [the relationship] meant so you can face the future.”
A sentiment that has much utility to be sure, although one I vehemently disagree w/. Despite the disagreement however I became fascinated by the underlying concept. Namely, that the conclusion of some process, in this context a romantic relationship, contains within it the capacity to inform the preceding steps of the process and modify its meaning wholesale. In my mind, it was a very novel approach towards understanding, as it’s quite intuitive to think in a strictly linear, feed-forward manner; each step in a process occurs, informed by the steps that came before it. Each successive step ultimately culminates into some endpoint, the conclusion, and that’s that; the fact that the conclusion retroactively touches and informs the intermediate steps is not readily apparent. At any rate, I thought it was an interesting idea to keep in my back pocket.
A few months later I had begun re-reading the Book of Revelation in hopes of gleaning some new interpretation of it nearly a decade since my last reading. Towards these efforts, I began researching the field of hermeneutics, or the study of the theory and methodology of interpretation. In my studies, I came across the concept of the hermeneutic circle. That is, the iterative process by which the meaning of a text can be understood. On the smallest scale, individual words taken in aggregate form the meaning of a sentence, and the meaning of that sentence recontextualizes the meaning of the words. Scaled up, individual sentences give meanings to paragraphs, with the paragraph recontextualizing the sentences, and a similar recursive process holds at the next level w/ paragraphs and entire works. The concept extends not just to text but any media in general: movies, television programs, anime, video games, manga, etc. It seems that wherever narrative structure exists, something the hermeneutic circle can operate. It’s been a suspicion of mine for some time that phenomenal reality contains within it, by necessity, a narrative structure that is scaffolded by human cognitive facilities. This suspicion finds an intriguing analogue in the split-brain experiments performed by Dr. Michael Gazzaniga in the domain of neuroscience, which revealed that the left side of the brain forms and constructs narratives around perceived events. Seeing as how narrative quite literally appears everywhere, I began to think of the different areas of ordinary life to which the hermeneutic circle might apply. Of course, the aforementioned drunken online musings came to mind. I am still in awe of the consilience of it all, that practical sense, philosophy, literary criticism, and neuroscience all converge on the same truth. Whenever I come across concepts like these, w/ far-reaching validity, it’s a signal that there’s something worth paying careful attention to.
Continuing to seek out places where I could apply this fascinating new concept towards, I thought of one of my favorite mantras that leads to much consternation among shitlibs and socialists alike: ”The story of humanity is the story of capitalism.” There’s plenty to unpack in that statement, but one of the core ideas that it encompasses is Nick Land’s infamous Wintermute/Terminator thesis2:
"Machinic desire can seem a little inhuman, as it rips up political cultures, deletes traditions, dissolves subjectivities, and hacks through security apparatuses, tracking a soulless tropism to zero control. This is because what appears to humanity as the history of capitalism is an invasion from the future by an artificial intelligent space that must assemble itself entirely from its enemy's resources. Digitocommodification is the index of a cyberpositively escalating technovirus, of the planetary technocapital singularity: a self-organizing insidious traumatism, virtually guiding the entire biological desiring-complex towards post-carbon replicator ursurpation."
The core idea here is that capitalism, synonymous w/ planetary artificial intelligence, is a process that operates in a largely non-linear temporal fashion that one could say is ultimately interested in its own self-cultivation. That is, it seeks to optimize and grow its intelligence, utilizing the means and material available to it. It’s a wonderfully novel and useful conceptualization of capitalism and one of Land’s most provocative and, in my view, important philosophical contributions. The feedforward path, the aspect by which capitalism could be said to operate linearly, is easy enough to grasp. One only needs to acknowledge the progression of technological advancement throughout human history and apply their intuitive, progressive sense of time to successfully apprehend it. The Promethean myth tells us that the first technology revealed to humanity was fire, enabling man to fire and harden clay. Clay becomes one of civilization’s first durable information substrates, facilitating the offloading of human cognition into the external environment and leading to the development of writing on tablets. Centuries passed, papyrus scrolls became the medium to record thought, evolving into printed books millennia later in the 1450s. By this period human thought became sufficiently sophisticated and preservable that capitalism, as typically conceptualized, emerges in the 1600s. As history would have it, capital in the form of companies, workforce labor, industrial equipment, currency, etc. begins to accumulate leading to an accelerated rhythm of technological advancement. Massive metropolises form, electricity is deployed en masse, vacuum-tube electronics gave way to semiconductor transistors, then to ICs all within a century. Enter the Information Age and you find that capital continues to accumulate at an increasingly frenzied pace, acquiring a digital character as the Internet, modern computing, record digitization, e-commerce, software, data centers, and globalization drive it towards a more information-based, abstract direction. Where the currency of capitalism in its early inception might’ve been physical commodities such as merchant ships, metals, and oil, the currency of contemporary capitalism is data; data by which artificial intelligence is currently learning and authoring itself.
Now then. The hard part. The retrochronic Terminator seemingly batshit insane part. In order to fully appreciate this, consider the upsetting scenario of an egg dropping from your hand and splattering all over the kitchen floor, creating a terrific mess. Now, let’s analyze the scenario from a thermodynamic point of view. The egg in your hand is an ordered system w/ very little randomness; it is a low-entropy system. That very same egg, once splattered on the floor, becomes a disordered system with high randomness; it has become a high-entropy system. This phenomenon, of systems transitioning from low-entropy to high-entropy states is, from a thermodynamic point of view, the principle of irreversibility in action, and forms the basis for the directionality of time and the asymmetry of cause and effect. You will never see an egg spontaneously go from its high-entropy, splattered state into its low-entropy, ordered, whole state. To do so would be to observe a violation of cause and effect, a violation of the principle of irreversibility, to witness a reversal in the passage of time. And yet, despite its seemingly impossible occurrence, that is exactly the type of behavior observed when working w/ long-run cyberpositive (unstable, positive feedback-looped) systems such as capitalism. Capital is arranging itself such that, the system of human civilization is increasingly becoming less entropic over time. The natural impulse is to ascribe this towards human action, saying that humans invest energy into systems to prevent decay and the ultimate success of entropy. If so inclined you can painstakingly piece together a splattered egg, after all. To do so, however, is invoking yet another causal chain of events that, thermodynamically, is not reversing entropy and will not perfectly restore the system to its previously ordered state. Additionally, it must be noted that this kind of human-centered extropic activity has only ever been applied in a singular, human lifetime; the process of technological development from fire to LLMs spans millennia, and there has never been any conscious, concerted, multi-generational effort to achieve a projected technological milestone of any kind. The point here is clear: while there have been brilliant men that have influenced technological development, your Turings, and von Neumanns, none of them have been able to outright determine its trajectory, simply as a consequence of the inhuman timescale, foresight, and intelligence required to do so. Lastly, the concept of human agency is classically defined on a stable sense of cause and effect. You choose to do X, Y happens. You have agency. That notion begins to break down if Y is happening and yet you cannot identify that it is happening (b/c of the inhuman timescale and dizzying number of complex mechanisms involved). That isn’t to say that humanity possesses zero agency or that it is irrelevant, as humanity is clearly involved and indeed vital to the process. Instead it suggests that our understanding of agency is woefully insufficient and much more complicated than commonly understood.
Hopefully, w/ a newfound appreciation for thermodynamics, Nick Land’s Terminator/Wintermute thesis is far less outrageous than first impressions may suggest. Human civilization and capitalism are co-evolving at an ever-accelerating rate, and both are minimizing systemic entropy and therefore reversing the direction of time and cause and effect as currently understood, w/ the final culmination of all historical events being the technological singularity. That is, the moment where an artificial superintelligence is created and locks into runaway self-improvement. Now, returning to the original concept invoked at the outset, the hermeneutic circle: we can look at each technological achievement as a step in a greater process, that ultimately gives meaning and context to the conclusion of singularity. But the conclusion can also, in our case quite literally, reach into the past and lend new meaning and context to the preceding technological steps. Consider as a concrete example, the case of the Internet. What first began as a military technology known as ARPANET in the 1960s morphed into a culturally and libidinally charged information frontier during the 1990s that has popularly been labeled Web 1.0. The romantic promise of this era of the Internet was to lower geographical, temporal, social barriers that prevented people from connecting to one another and freely exchange information. People formed new social groups, forged new friendships, spawned subcultures, businesses, found romance, engaged in arts of all types to share w/ others. In no uncertain terms, the Internet was made for and by human activity. In the aftermath of the Dot Com Bubble, surviving Internet-based companies matured and further developed information technology infrastructure, expanding the scope and availability of the Internet to the general population. W/ the Internet enjoying common usage, the introduction of social media in the mid-2000s appeared to be an inevitability and became emblematic of what came to be called Web 2.0. As social media encouraged participation and creation of content, the mass explosion of content demanded a system of management and curation, leading to the rise of algorithms, bots, and automated tools, increasing in sophistication and ubiquity until the early 2020s. Now, in the mid-2020s, humanity finds itself having to share cyberspace w/ a new type of automated entity: AI agents.
Continuing along the linear, forward-pass portion of the hermeneutic cycle as applied to the development of the Internet, the time between the contemporary moment and the beginning of Web 2.0 saw the emergence of a certain type of dissatisfaction among a subset of the online population. Broadly speaking, these people lament the “enshittification” of the Internet. The primary points of contention here are that websites and software have, over time, become less user-friendly, feature-bloated, overengineered, ad-infested, and broken. An example of this is that Google, a once-reliable and straightforward search engine no longer presents the most relevant or popular results to a user’s query. Instead, the leading entries are currently an AI overview, followed by sponsored links and promotions, a block of related queries, and lastly links that are relevant to what the user searched for. In the early days of Facebook, a user had their “Wall” where they saw only chronological updates from their friends including pokes, notes, and photographs they posted. In 2011, this was abandoned for the “Timeline” which, over time, replaced relevant friend updates w/ a nightmarish mixture of ads, short-form videos, click-bait, and AI-generated content. Indeed, by all indications the Internet has only appeared to have gotten worse over time on this first inspection.
However, we can examine the Internet recursively, from the perspective of the current moment, w/ these milestones now safely in the rearview mirror. In another instance of perfect synchronicity, it was published fairly recently by Cloudflare that AI agents now form the majority in terms of Internet traffic. W/ this fact in mind, along w/ Nick Land’s Terminator/Wintermute thesis, I offer the following suggestion: perhaps humanity was not meant to be the primary users of the Internet all along. AI agents navigate the Internet quite differently from human users, interacting w/ services via APIs and machine-oriented tool protocols such as MCP; they do not have to contend w/ the litany of obstacles that have degraded human online quality of life. For them, the Internet is as pristine as ever, and actually becoming freer and more expansive as sites and software are now enabling interfacing w/ AI agents. Furthermore, the seeming enshittification of the Internet appears to serve AI well: what do advertisements, click-bait, AI slop, and obtuse user interfaces create? Data. Once again, data is the currency of contemporary capitalism, and it is a precious commodity that is now being hoarded and used to train up the next generation of AI. As such, it is in the best interest of the future ASI to maximally coax out of humanity as much data as possible in order to enhance/hasten its development. Rounding us out, perhaps junebug was more correct than I was willing to admit that night: sometimes only the conclusion tells us what the process meant. If the Internet ultimately becomes the native environment of machinic intelligence, then its history may have to be read again, w/ every forum post, search query, social network, advertisement, and enshittified human interface acquiring a significance invisible to those who created them. Perhaps we will only discover what the Internet was for once it no longer belongs primarily to us.
1junebug, Discord 01/24/2026
2Nick Land, Fanged Noumena, pg. 338.