Epstein and Nowak were consumed by the trajectory of the human species and by what they perceived as the catastrophic implications of “reverse Darwinism” — the proposition that as technological civilisation advances, the biological substrate of the species deteriorates. They did not treat this as an abstract intellectual worry. They treated it as a problem requiring intervention. At a breakfast gathering in Epstein’s 50,000-square-foot Manhattan residence — attended by former heads of state, senior executives from Google, and at least one former Israeli prime minister — Epstein articulated Nowak’s research for his distinguished guests. The manner in which he framed it was diagnostic. He explained that when one seeks to identify a suspected malignant actor on the global stage, one gathers intelligence — the signals, the connections, the relational web — and then one eliminates the actor by severing those connections. He was ostensibly describing cancer cells. But he was not, in any meaningful sense, describing cancer cells. In conversations with numerous high-level interlocutors, including Nowak himself, Epstein had an unmistakable habit of deploying biological metaphors as euphemisms for something far less clinical. He spoke of the human species as one might speak of a bacterial culture. Dead organisms, he would say, must be washed away. Those that are no longer useful to the body — or to the system — those that are cancerous — they need to be eliminated. When he spoke of cancer, he was not speaking of oncology. He was speaking of human beings he deemed surplus.
Epstein, for his part, was fixated on cryogenically preserving his own reproductive material. He wished to perpetuate himself. The eugenicist orientation was not subtext. It was not metaphor. It was the declared operating philosophy of a man who funded the laboratories, cultivated the scientists, and socialised with the policy architects who shared his conviction that democracy was a failed experiment, that humanity was barreling toward obsolescence, and that the emergence of artificial super-intelligence would render most of the species redundant. And it is at this juncture — where the ideology becomes operational — that the truly dangerous thinking commences: the notion that the human race must be “tuned” into something more sustainable, even if “sustainable” does not encompass all of us. These are not novel ideas. They are the same ideas that animated the eugenics programs of the early twentieth century, the same ideas that Adolf Hitler attributed to Friedrich Nietzsche, the same ideas that have resurfaced, generation after generation, whenever a sufficiently wealthy and sufficiently insulated cohort convinces itself that the mass of humanity is a problem to be managed rather than a constituency to be acknowledged & served. But Epstein’s fascination with AI and his fascination with eugenics were not parallel interests. They were convergent aspects of a single unified project. And to comprehend how that project connects to the device in your hand — every time you query ChatGPT, scroll through Facebook, navigate LinkedIn, surrender your location data to Spotify — you must understand the relevance of a DARPA initiative called LifeLog and the strategic calculus of Peter Thiel.
DARPA — the CIA-affiliated research apparatus responsible for incubating technologies that range from the internet itself to, as will become apparent, the social media platforms you inhabit — operated a program called LifeLog beginning in 2003. The original request for proposal remains publicly accessible. The program’s architecture was explicit: capture the totality of every American citizen’s informational footprint, aggregate it into a comprehensive longitudinal record, and construct from that record an electronic database encompassing effectively every individual on Earth. When congressional oversight exposed LifeLog’s existence and the American public recoiled at the implications — this was 2003, an era when people still retained a residual wariness about what they surrendered to computers, when dial-up modems and AOL were recent memory — DARPA terminated the program. The termination date was February 4, 2004. That date is indelible because it is the precise date on which a twenty-year-old Mark Zuckerberg, seated in a Harvard University dormitory, inaugurated a website he called “the Facebook.” Facebook was conceived as nothing more than a campus novelty. Zuckerberg, by his own account, intended it as a Harvard-internal project and nothing beyond. He had encountered the Winklevoss brothers — Cameron and Tyler, affluent and well-connected — who collaborated on its early development. The popular mythology, as dramatised in a feature film, holds that Zuckerberg pivoted away from the Winklevoss vision and built something revolutionary on his own. This is not the story the evidence tells. The actual story begins with what happened immediately after launch. Facebook acquired tens of thousands of users. Half of them were on the Harvard campus. It faltered at other universities. Growth was anemic. The platform was going nowhere. Then Peter Thiel materialised. Thiel — co-founder of PayPal, architect of Palantir Technologies, and today one of the most influential political donors in conservative American politics — approached this unremarkable college student and extended $500,000. Thiel was not, in any conventional sense, an angel investor. He was an ideological recruiter who paid promising young people to abandon their educations and subordinate themselves to his enterprises. His network had birthed or catalysed LinkedIn, Instagram, Spotify, and scores of other platforms. The question that demands an answer is elementary: why would a man of Thiel’s sophistication, strategic depth, and access to virtually any investment opportunity on the planet commit half a million dollars to a struggling dormitory experiment with no discernible revenue model?
Because of Palantir.
Thiel had founded Palantir Technologies as a data analytics engine — a system designed to ingest enormous volumes of disparate information, identify latent connections between data points, and synthesise conclusions that no individual analyst could reach unaided. The technology was formidable. Its analytical capabilities were, and remain, extraordinary. But Palantir had a fundamental dependency: it required data. Humungously large, continuously replenished quantities of human behavioural data. Location. Social connections. Consumption patterns. Emotional reactions. Ideological affiliations. Palantir was the brain. But a brain starved of sensory input is inert. LifeLog was supposed to be the feeding mechanism — a government-administered pipeline that would deliver the informational lifeblood of every American citizen directly into the analytical apparatus. When the public killed LifeLog, the pipeline had to be rerouted. It was rerouted through social media. Through platforms that people would not merely tolerate but embrace, celebrate, defend, and integrate so deeply into their daily existence that the idea of living without them would come to feel like amputation. John Poindexter — the former National Security Advisor who had directed DARPA’s Information Awareness Office and conceived the Total Information Awareness initiative, the explicit progenitor of mass domestic surveillance — met with Thiel and Palantir co-founder Alex Karp in 2004, according to Wired magazine. Poindexter proposed applying data-mining methodologies originally developed for financial fraud detection to counterterrorism operations. The post-September 11 legislative environment — the PATRIOT Act, the FISA courts, the sweeping surveillance authorities signed into law by President Bush — had vested the intelligence community with historically unprecedented powers of information collection. But the decisive breakthrough was not legislative. It was psychological. Americans would not countenance a government program that harvested their data. They would, however, surrender that identical data — voluntarily, copiously, compulsively — to a free application that permitted them to share photographs of their lunch, pets, children and also quarrel with distant relatives about politics. Two decades later, CIA officials would acknowledge, with the understated candour characteristic of people who have already won, that platforms like Facebook, LinkedIn, and Spotify had become the single most productive source of intelligence on the American civilian population.
The inspiring narrative of Mark Zuckerberg — the prodigy who built an empire from nothing — was never what it appeared. Virtually every dominant social media platform of that era received early capital from investors with profound, documented connections to DARPA and the intelligence establishment. MySpace lacked that patronage. It perished. Facebook had it. It conquered. The pattern replicated itself across the entire topology of digital life. Reid Hoffman — LinkedIn’s founder, another filament in this web — hosted a dinner in Palo Alto that seated Epstein at the same table as Musk, Zuckerberg, and Thiel. Each of them would subsequently become an evangelist for cryptocurrency — because once the informational architecture is operational, the logical subsequent manoeuvre is to separate the financial system from governmental oversight entirely. Deleveraging finance from the state would enable the technocratic class to consolidate the one form of power that remained partially beyond its grasp. Bitcoin and its progeny were not speculative curiosities. They were instruments for constructing a parallel monetary infrastructure — one in which the custodians of the data could simultaneously become the custodians of the capital flowing through it. The vision was architecturally coherent: commandeer the data, dominate the analytics, fabricate the financial rails, and eventually render the nation-state superfluous. You do not overthrow the government. You make the government irrelevant. Then Edward Snowden intervened. In 2013, Snowden — a man one might once have condemned but whom history may ultimately vindicate — disclosed the existence of PRISM, a surveillance programme originating from software appropriated by the government during the Clinton administration. PRISM possessed the capacity to aggregate informational profiles on the entirety of the American populace and to harvest intelligence from overseas communications with equal facility. Snowden was a whistleblower in the most consequential sense of the term: he demonstrated, with documentary proof, that the CIA and the NSA possessed the operational capability to monitor every American citizen through every digital device they owned. And that is where Facebook data was being routed. There exists a legal prohibition against the American government conducting surveillance on its own citizenry. But when citizens volunteer their information — when they offer it up eagerly, gratuitously, in exchange for the dopamine reward of a notification — all the state requires is a backdoor. That is why it was imperative for Thiel and Poindexter to establish these operations. It is also why both Donald Trump and Joe Biden — in a rare display of unanimous bipartisan urgency — pursued the curtailment of TikTok’s presence in the United States. If your government possesses backdoor access to Facebook and Google, and a foreign adversary does not, the asymmetry favours you. When a Chinese-controlled application begins harvesting identical volumes of data on American citizens — data to which your intelligence services have no access — the asymmetry inverts. The campaign to eliminate TikTok was arguably the most transparent admission in recent memory that the relationship between the U.S. government and Silicon Valley is not regulatory. It is operational……………….
(Ed. For previous episodes in the series ‘click’ on ‘DJM’ below.
© DJM 2026
