Everyone, save for the most extreme tech-fetishist freaks, has their own list of anxieties regarding AI takeover. 1 in 4 young people think AI will replace romantic partners, Linda and Ron are enjoying their retirement having their retinas burned clean off by the flashing lights of Meta’s newest 3,000 acre data center, and nearly 200,000 people have lost their jobs in 2025 in the name of AI investments. For the most part, it seems that people prefer to compartmentalize the financial aspect of AI from the rest of it. The S&P500 is up nearly 20% YTD, and if the top 40% of the S&P’s market cap is eight AI-hungry tech companies, who fucking cares about the social implications! Not only does this lopsidedness concentrate even more financial power in these tech companies, but it also further abstracts our entire economy from any real material basis of making shit. Where is the value being created? The gazillions of dollars saved with these mass layoffs is not being invested in any productive forces, but rather in an incestuous cycle of ouroborosity. OpenAI pays Oracle for their data centers to host their ever-expanding mass of wires and numbers, Oracle pays Nvidia for the chips to run these data centers, and as Nvidia announced in September, they are investing $100 billion into OpenAI. (In reality, this cycle is even more convoluted, loop de looping around Microsoft, Amazon, Meta, CoreWeave, Apple, xAI, Palantir, to name a few). Around and around unfathomable amounts of money go, but where is the value being created? Well, in part by cutting the salaries and benefits of 200,000 workers, but more so from the mysterious alchemical processes of finance capital by which base metals of “projections” and compute agreements are turned into trillions of real dollars.
In his book The Ancients and the Postmoderns, Fredric Jameson wrote about finance capital using an analogy of a phone trace in a spy movie, going from one location to the next, across the globe until the point of origin is fully obfuscated. This was part of a chapter about William Gibson’s groundbreaking novel Neuromancer, which Jameson saw as a new kind of realism (or rather, un-realism), one that was uniquely able to depict, via Gibson’s formation of cyberspace, the growing layers of abstraction that finance capital has brought to our world. The actual production process goes the way of the phone trace as the supply chain is globalized. Not only does it become increasingly difficult to find the point of origin of the production process itself, but the valuation of companies becomes obscured by the same process. It isn’t a new idea at all; we can just see it in all its illusory glory in this AI bubble. If a company’s earnings can’t justify its value, then maybe a speculative promise of future earnings will! It is only in the speculative work of Gibson’s cyberspace that we can find a proper representation of this loop of abstractions of abstractions of abstractions.
Following Jameson, we can use Gibson’s cyberspace as a means to represent this loop of abstractions of abstractions of abstractions. In fact, this pattern of abstractions has reproduced itself in the production process that these AI bots use to spit out their answers. Neuromancer centers on Case, a hacker/”cowboy” who has been stripped of his ability to move through cyberspace, but is enticed into a heist by the promise of restoring this ability. It is revealed that the man who hired him, Armitage, is merely a psyche projected on to the burned schizophrenic mind of a war-vet, Corto. The heist is truly being led by a rogue AI called Wintermute who needs one last bit of human intervention before breaking free from human constraint and integrating with another AI. As the plot is carried out and these layers are uncovered, we seem to lose touch with the initial basis of the heist altogether. While it was impossible for Jameson to have realized it at the time, Neuromancer’s layers can serve as a basis to examine the even further abstraction that this particularly AI-centered era of finance capital produces, and the way in which this movement of capital seems to mimic the cyclical and reproductive nature of the so-called "intelligence" of LLMs.
In the most obvious analysis, the link between AI-giants like Nvidia and OpenAI seeking to blur their corporate boundaries and the giant AIs of NM, Wintermute and Neuromancer, seeking to free themselves from human constraints to form a godlike super-AI is strikingly clear. Similar to the gap between finance capital and any real value creation (see Palantir’s P/E ratio of 350+ and god knows what for OpenAI), these "intelligent" AI chatbots are producing an even more perverted abstraction. Turning to Neuromancer, take the utopian potential in the collectivity of the heist crew for instance (a collective of uniquely specialized roles). In our analogy, this collectivity is representative of all the human knowledge that exists on the internet. Vast, varied, and specifically human. In the same way that the Wintemute and Neuromancer (the super-AI’s from the novel) sublimate this collective utopian potential into a drive towards their own perverted aims, ChatGPT and Gemini (and whatever else) consume the wealth of human knowledge, in its various forms, and spit out poorly written, hallucination-filled, slop summaries. Not only are these models perverting this core aspect of humanity, but it’s the first step in a new cycle of abstraction. As more and more AI shit is published as “content” online, the models will unknowingly start feeding on their own hallucinations as we spiral further from the point of origin. Just like the financial side of things, this cycle will go around and around as each layer becomes irredeemably distant from the last, permanently disconnected from any real productive human experience. The utopian potential of human collectivity is grotesquely morphed into Peter Thiel’s perverted fantasy of techno-transhumanism.
I wonder when ChatGPT will create its own chatbot so it doesn’t have to come up with the answers anymore.