How Huxley's 1961 'Law of Technology' Became the Algorithm

Opening

Dear reader, I recently watched a 1961 BBC interview—an 8-minute conversation between Aldous Huxley and host John Morgan. It’s a black-and-white recording more than 60 years old, yet what the author said in it described 2025 with an accuracy that genuinely gave me chills.

Huxley put it this way: “If you plant the seed of a technology, it will grow according to its own laws of being—and those laws aren’t necessarily the same as our human laws.” He added: “It is possible to make people contented with their servitude.”

Bread and circuses, endless entertainment and propaganda—this is exactly what’s happening inside our smartphones every single day. Today, I want to talk about how Huxley’s warning is playing out in the age of AI, and what “solution” he ultimately left us with.

Technology Grows by Its Own Laws

Appearing on BBC’s In Conversation series on July 30, 1961, Huxley was blunt: this wasn’t a problem specific to any one nation. Whether in the United States, Britain, or Russia—once a technological means is developed, it will eventually be used. Technology grows according to the logic of its own existence, regardless of human intention.

This idea wasn’t Huxley’s alone. Around the same time, the French sociologist Jacques Ellul¹​ made a nearly identical argument in The Technological Society (1954). Through his concept of “technique,” Ellul explained that technology isn’t a mere tool but a system that autonomously expands toward ever-greater efficiency. Interestingly, Huxley directly references Ellul in this very interview, calling him “one of the most competent writers on sociological matters.”

Both men arrived at the same core insight: technology is not a neutral tool. Once it exists, it expands according to its own logic, and in that process, humans become subjects rather than controllers. Huxley called this “the Frankenstein monster problem.”

In 2025, this “law of its own” is operating most vividly in the form of AI recommendation algorithms. Social media’s AI recommendation systems are designed to maximize how long users stay on the platform. According to MIT Technology Review, these systems analyze micro-level behavioral data—how long a user lingers on a particular image, how often they check notifications—to construct a personalized “dopamine schedule.” The algorithm isn’t understanding the user. The algorithm is autonomously evolving toward its own goal (maximizing time spent).

Huxley’s claim that “technology grows according to its own laws of being” has, 60 years later, been literally realized in the form of reinforcement learning²​ models.