The slow death of the accelerationist.

The year is 2024. Summer has just begun. National discourse, for now, is solely focused on the upcoming presidential election, with many a journalist or political commentator critiquing the current, rather fiery state of political affairs. Tech and its associated public commentary has centered upon artificial intelligence as its new darling, hailing OpenAI as  a savior for what was once deemed an idea stuck in science fiction, and looking to burgeoning startups such as Cursor and Windsurf as early examples of how agents could automate software engineering tasks. Logging onto Twitter, one would catch glimpses of Beff Jezos, an aptly named satirical account, relentlessly posting optimistic odes about how our own silicon creations will soon enable us to solve all of our problems, enabling us to truly accelerate. Beff's social posts were not just the isolated ramblings of an overtly verbose anon; they were slowly becoming a zeitgeist of their own, inspiring an entire independent cohort of individuals who slowly began appending their public profiles with 4 characters: e/acc.

The e/acc community was, despite its overarching and centric belief, surprisingly diverse. You could find bootstrapped startup founders working on their next B2B SaaS play, far right exhibitionists who were enjoying both the attention and money that Twitter's creator program had bestowed upon them, and renowned venture capitalists, all espousing the same ideas. One could argue that the central tenet of e/acc was rebellion: rebellion against the status quo, rebellion against the government (or the broader powers that be), and rebellion against those who may have doubted them in the past. This haphazard group slowly began to gain momentum, with Beff Jezos, who was later doxxed and revealed to be former Google scientist Guillaume Verdon, creating his own hardware startup, Extropic.

The year is now 2026. The e/acc movement is now, for the most part, dead, with little to no mention of it on Twitter, or on popular technology podcasts. The remnants of the community no longer sing praises for a technology that is still yet to come; they instead attempt to convince each other that their application of said technology are morally, ethically, or technically superior than that of the anonymous Discord user typing below them.

The history of AI, albeit short, is already incredibly rich. Never before has a certain technology changed this quickly, and brought with such rapid alterations in how we perceive the world, and ourselves. The summer of 2024 remains an interesting and somewhat unique time in this history: ChatGPT and its counterparts had been around long enough to become a part of the public discourse, but yet were still close enough to their infancy that it was not quite certain what they could become, or where the technology would eventually go. This effect was felt across the social, economical, and philosophical extensions of the colloquial world of "tech", which at that time, seemed to be all-encompassing: indeed, it might be years until we understand the extent to which this particular circle of individuals had an effect on cultural norms, politics, and more during this period, largely as a result of the optimism in which everyone felt at the time. AI is no longer an optimistic technology. As with any new technology, the honeymoon period has effectively ended. The same university students who raved about the latest release of GPT to their classmates are now dreading the prospect of entering a job market that is both challenging and ever-changing. The same tech bros who were early to vibe-coding are now lamenting the loss of technical moat for their businesses. The looming threat of economic risk, a risk that was once dismissed as hearsay and doomerism by those in the techno-bubble, is now very real. The national pride that once accompanied the advancement of AI being solely in the hands of American-made startups has evaporated, with Chinese counterparts such as DeepSeek and MiniMax shipping equally capable, open-source counterparts at a fraction of the price.

As we continue to grapple with and lament the changes that AI has brought us, I am often reminded of some conversations I had with friends who were around for the early days of the internet, back when AOL was the primary messaging app, and back when you could apparently find early drafts of internet-based currencies that predated Bitcoin. The internet at that time was, to many, special. Being a hacker, a person who knew their way around computers, networks, and the like, was a social boon, not a black mark. But yet, as the internet evolved, as it became commoditized and invaded by corporate whims and infrastructure, it became plain, a tool that enhanced productivity, but did little else for the soul. Being a hacker meant being embroiled in controversy or criminality, or worse, being a social outcast or nerd who could barely hold a conversation with their fellow man. The internet had produced an identity, one which got lost and eventually cast aside once the underlying technology became commoditized. Even within the subset of the internet that still considered themselves true hackers, there were now various gatekeepers, gatekeepers whose standards you had to meet before you could publicly proclaim yourself as a member of the broader collective of the hacker community. And with that, "hackerism" went from a cultural-norm, back to colloquial term associated with men in dark rooms, wearing black jackets and typing away at a neon keyboard. A movement can weaken at the very moment its central object becomes more important, because what made the movement compelling was never just belief in the object’s importance. It was the sense that belief itself distinguished you. Once everyone agrees that the technology matters, the movement loses one of its primary functions.

A similar thing happened to accelerationism. The technology movement of the time, AI, came by, became special, and then became mainstream, but this time, with nothing to replace it. Popular sentiment went from blissful glee to unabashed debate, debate on if the environmental costs of developing better AI models was worth, debate on if the economic uncertainty caused by increasingly autonomous models would become more severe. The market was flooded by a wave of startups building AI-based tools for a plethora of use-cases. It is difficult to sustain a politics of unbounded technological optimism once the technology in question no longer feels singular. It is difficult to maintain the romance of acceleration when what acceleration mostly seems to produce is an endless stream of mediocre products, collapsing defensibility, and a strange sense that capability is everywhere while meaningful progress remains harder to locate than expected.

And that, more than any technical disappointment, is what the accelerationist could not survive. While the individual downturns of more recent movements, such as the hackers, the NFT shills, and the toxic masculinity stans was due to our broader potopurri of culture either rejecting them or their movements failing due to economic or social pressures outside of their control, AI accelerationists have neither assimilated, nor have they been rejected. They have simply been left to be, left to wallow in an ironic reality in which their special technology progressed at a pace faster than anyone could have hoped, but yet, it became known not for enabling unprecedented societal progress, but for becoming a part of the stack, the same stack of software aided productivity that society slowly began to accept as a norm, a norm that became more associated with its negatives in public opinion than its positives, just like social media and cryptocurrency before it.

The summer of 2024 may remain a small footnote, if that, in the broader history of the development of AI. Yet, for those that were actively, for the lack of a better term, "chronically online", it may represent the last peak of the accelerationists, the tech bros, the culture of builders. While past trends such as the dot com bubble and social media applications had created similar microcosms of closed-off cults that had similarly either died off or assimilated into a wider societal group, the progress of AI is different altogether. Accelerationism did not get proven wrong (AI hype is at an all time high) nor did it fizzle out: it simply became normal. Being an optimistic accelerationist is fruitless when the technology you are interacting with is no longer special, no longer a science-fiction dream come to reality. It is not a badge of honor when it feels that everyone can solve or do anything, yet nothing is actually getting done. As the umpteenth vibe-coded app hits the market, it is worth wondering what happened to the collective optimism of the tech community just a year and a half ago. For as it stands today, it seems that we are currently living through not unbounded accelerationism, but rather the slow death of the accelerationist.