Prophecy as a Service: Leopold Aschenbrenner's 165-Page Pitch Deck
Sprawling AI manifestos are not visionary roadmaps, but desperate marketing brochures meant to recruit exit liquidity for an over-leveraged hardware bubble.
By Ida Vann
Sparked by Situational Awareness and the Impending Stock Market Volatility · discussion
“AGI by 2027 is strikingly plausible,” declares Leopold Aschenbrenner at the start of his 165-page manifesto on the destiny of human intelligence. I spent the better part of my afternoon reading this sprawling treatise, groaning at the sheer density of the philosophical epic warning of superintelligence, geopolitical doom, and the inevitable dawn of machine gods. I have waded through a massive volume of bloated Silicon Valley treatises in my time, but it takes a truly special kind of hubris to publish a 165-page prophetic manuscript, embark on an exhaustive podcast media blitz to warn the public, and then immediately reveal that the next logical step in securing humanity's future is launching an investment firm focused on AGI, with anchor investments from Patrick Collison. We must abandon the habit of treating these sprawling artificial intelligence manifestos as visionary technological roadmaps, and start reading them as desperate marketing brochures meant to recruit exit liquidity for the authors' own portfolios.
If you dig into the actual PDF document, the forensic reality of the text is staggering in its presumptions. Aschenbrenner dedicates entire chapters to mapping out the creation of trillion-dollar compute clusters, casually tossing around predictions that the United States government will soon need to nationalize data centers. He asks the reader to accept, as a baseline, that within three years we will have autonomous systems capable of triggering a runaway intelligence explosion. It is written with the grave, burdened tone of a physicist warning the president about the Manhattan Project. But Robert Oppenheimer didn't follow up his security clearance briefings by opening a boutique wealth management firm.
This brings us directly to the underlying financial reality of the pitch deck. If you ask the proof-of-superintelligence crowd, the stakes here are strictly existential, and they have some undeniably compelling surface-level arguments regarding national security and the theoretical limits of hardware scaling. I will happily concede that global technological supremacy and the race to secure semiconductor supply chains are genuine geopolitical concerns. But Aschenbrenner’s pitch itself rapidly abandons this high-minded philosophical framing the moment it hits the open market, translating all that speculative sci-fi rhetoric into a remarkably mundane financial vehicle. Strip away the dense jargon about intelligence explosions and temporal alignment, and what you are left with is an unhedged, over-leveraged proxy for a semiconductor ETF.
To fully grasp the absurdity of this transition, you have to look closely at how the machinery actually operates behind the curtain. The author posits a world where multi-trillion-dollar server clusters are an absolute certainty by the end of the decade, demanding an unprecedented, civilization-scale capital formation to survive. The proposed solution to this looming apocalypse is a sprawling clusterfuck of napkin math where investors are urged to pour massive capital into compute-adjacent equities today. If you aren't familiar with how these compute clusters are financed, it largely boils down to a massive circular economy: cloud providers dump billions into buzzy AI startups, on the strict condition that those startups immediately hand the money right back to the cloud providers to rent server space.
The ecosystem surrounding these predictions has become so completely detached from empirical reality that industry observers are now forced to treat sci-fi roleplay as rigorous financial planning. Earlier this week, I was reviewing a speculative forecasting exercise by Wojciech Gryc modeling the downstream macroeconomic effects of this exact flavor of AGI betting. Analysts are already projecting disastrous 2026 margin calls based on the widespread market adoption of these tech-utopian investment strategies. When an entire financial sector begins treating a 165-page philosophical thought experiment as binding forward-looking earnings guidance, the traditional guardrails of risk management have completely evaporated. Investors funneling cash into these funds are simply buying equity in a prophecy, ensuring the mechanics of the actual miracle remain conveniently obfuscated behind demands for ever-increasing capital expenditure.
We have seen this precise structural architecture before. The current artificial intelligence market’s leveraged margin loops are structurally identical to the 2022 crypto crashes that evaporated billions of dollars of retail wealth. During the collapses of Terra, Celsius, and FTX, the systemic rot was driven entirely by a self-referential cycle of tokens backing other tokens to artificially inflate yields, all disguised beneath the utopian rhetoric of decentralized finance and banking the unbanked.
Today, the artificial intelligence industry is running the exact same playbook with slightly more sophisticated vocabulary. Major firms are borrowing heavily against hypothetical future compute capabilities to purchase current compute hardware, thereby pumping the immediate market valuations of the very compute providers they rely upon to secure further debt. This is a classic death spiral mechanism dressed up in a Patagonia vest. Aschenbrenner's newly announced fund is just the current cycle's preferred vehicle to privatize the upside of this speculative loop while ensuring the broader market is positioned to socialize the inevitable collapse. The collateral backing these massive systemic bets consists entirely of the raw, unadulterated belief that a superintelligent god will magically emerge from the server racks in thirty-six months to pay off the interest, completely untethered from any verifiable software revenue or enterprise adoption.
The brutal dissonance between the cosmic stakes of the manifesto and the grimy reality of the fund structure is impossible to ignore when examining how actual technologists are reacting. You can toggle directly from the manifesto's breathless warnings about humanity's final invention over to a remarkably grounded Hacker News comment thread dissecting the pure financial grift of the maneuver. The engineers and developers tasked with building actual software are overwhelmingly pointing out the sheer marketing audacity of treating an ETF-proxy launch as a world-historic philosophical event. The author spent 165 dense pages arguing that we are hurdling toward an intelligence explosion that will fundamentally rewrite the laws of physics and human society, and his immediate, actionable response to this looming singularity was to pool capital from the Stripe guys to collect a standard two-and-twenty management fee.
That would be a hedge fund.
We’ve already seen what happens when we let a bunch of over-leveraged techno-utopianists use pseudo-intellectual manifestos to gamble with the broader economy. We cannot afford the apathy of assuming these isolated Silicon Valley ecosystem bets will remain safely quarantined to the venture capital class. The fallout from a trillion-dollar hardware bubble bursting will not be contained to Sand Hill Road; it will bleed into public pension funds, retail retirement accounts, and the broader hardware supply chain. It is time to demand rigorous documentation of AI leverage, strip away the sci-fi marketing, and force these firms to operate strictly within the bounds of material reality. Because caring about the next financial contagion isn't naive, and we shouldn't wait for the multi-billion-dollar margin calls to start demanding basic regulatory accountability. We can simply choose to stop treating their pitch decks like holy texts.