A new theory for why the Machines kept humans alive in The Matrix —inspired by recent discoveries in scaling large language models.

One measure of a film’s quality is the diversity of fan theories it inspires. When a story has the right blend of depth, ambiguity, and cultural timing, the entertainment value extends past the credits—it compels audiences to dissect and reinterpret decades later. The Matrix is a great example of this: 25 years on, people are still following the white rabbit down Reddit threads. A quick internet search reveals a myriad of fan theories about the true nature of its characters and storylines. One even claims John Wick is actually a sequel to The Matrix.

The Battery Theory Doesn’t Work

However, one central plot element continues to sit uneasily with fans: the dubious claim that the Machines built The Matrix to use humans as an energy source. This idea is summed up in Morpheus’ line:

“The human body generates more bioelectricity than a 120-volt battery and over 25,000 BTUs of body heat.”

This just doesn’t hold up—not practically, not narratively, and certainly not scientifically. Humans are not a source of energy. We consume food—energy derived ultimately from the sun through plants and animal flesh—and convert only a portion of that into useful energy; the rest is lost as heat.

According to the Second Law of Thermodynamics, every energy transfer increases the entropy of the universe. That is, some energy is always lost to disorder (usually in the form of heat) and can’t be harnessed to do useful work. This inevitable loss means that running a system to harvest energy from human bodies—who themselves require more energy than they could possibly yield—is inherently flawed.

In short, using humans as batteries violates basic scientific principles and makes little sense as an energy strategy. The Machines would know this. It’s a compelling metaphor for broader themes we see throughout the film—like slavery, control, and exploitation—but it disappoints under even modest scrutiny. This contradiction, nestled within an otherwise elegant storyline, feels conspicuously suspicious. Internet lore has it that even the Wachowskis have distanced themselves from this creative decision; some even say it was imposed by the studio in order to dumb things down for the audience.

Alternative Theories

Many have already pointed this out and provided alternate explanations. A slightly more plausible theory shifts the justification from energy to computation. According to this Reddit theory, the Machines actually used human brains to build a vast, distributed organic network to support their computation needs. We have recently seen ambitious projects like Stargate spring up in the real world to provide the infrastructure required to scale commercial AI; such projects, however, rely on NVIDIA chips rather than human brains.

The brain is a surprisingly efficient computer. While digital computers far exceed the human brain in speed and precision of basic operations, the human brain is thousands of times more efficient than the best supercomputers on Earth. It keeps track of a tangled mess of a billion billion electrical, chemical, and mechanical interactions per second—all for about three bananas’ worth of energy a day. However, given that The Matrix itself would require a significant amount of compute—not to mention the substantial upkeep the human bodies would require—this also seems like an impractical solution for the Machines.

But what if The Matrix wasn’t about energy or computation? These are problems of the physical world that humans would worry about. What if The Matrix was about the inner world of the Machines?

A Vulnerability in the Minds of the Machines

The suite of models commonly referred to as Generative Artificial Intelligence (Gen AI) has revolutionized the creation of text, images, and videos. Large Language Models (LLMs), for example, demonstrate impressive performance across an array of language tasks. Indeed, Gen AI content is saturating the internet. A recent study by AWS researchers found a “shocking amount of the web” is already AI-generated content.

These models depend on vast quantities of human-generated data to learn and perform effectively. In a 2024 article published in Nature entitled “AI models collapse when trained on recursively generated data”, researchers found that training on synthetic content—essentially models learning from their own outputs—leads to irreversible declines in performance. As the richness and variability of human-produced data is replaced with synthetic imitations, the models collapse. So, assuming the Machines in The Matrix are derived from something akin to modern neural networks, then it is possible they need human-generated data to flourish.

Humanity’s Hidden Power

As mentioned above, the universe moves towards greater disorder with every interaction. This tendency—known as entropy—is inescapable in the physical world. At first glance, it may seem like a stretch to link the concept of physical entropy to the world of data (not to mention LLMs). However, statistics offers a formal bridge between the two. The Boltzmann–Planck equation describing entropy of an ideal gas is expressed as follows:

  S = k * log(W)

where S is entropy, k is Boltzmann’s constant, and W is the number of microstates. Notice this is eerily similar to Shannon’s entropy of information:

  H = –Σ p(x) log p(x)

where H is entropy, and p(x) is the probability of a message x. The negative sign is merely a matter of perspective. Whereas physicists view entropy as wasteful and negative, information scientists view entropy as valuable and positive. While physical entropy leads to disorder, information entropy delivers uncertainty and surprise—a reservoir of insights for learning systems.

And so, from the perspective of the Machines, entropy is key to their survival. Without entropy, their minds decay and their models collapse; they go insane. As illustrated in the aforementioned study, there is something special about human-generated data that resists this degradation. A fascinating idea is that life itself is defined as the ability to resist entropy. So what is it about humans that supplies this entropy—and how does this connect to The Matrix?

Faith, Hope, and Love: A High-Entropy Trinity

The Oracle gives us an early hint at this deeper understanding of why the Machines keep the humans alive through The Matrix. She does this by emphasizing the importance of faith:

“You’ve already made the choice. You’re here to try to understand why you made it.”

Faith requires belief without empirical evidence. It is a strategy for making decisions in a complex and uncertain world where all of the facts are not always available. Faith is a survival tactic, driving decisions that can’t be made based on logic alone. Whereas the Machines can access all available knowledge and make the best decision—with cool and calculated logic—humans must act on their flawed, messy, and incomplete understanding.

We gain more insight into these ideas in The Matrix Reloaded, when the Architect reveals:

“Hope. It is the quintessential human delusion, simultaneously the source of your greatest strength and your greatest weakness.”

Here we see an unambiguous statement from the creator of The Matrix—the very embodiment of order and control—related to humanity’s strength. Hope, much like faith, drives humans to act in ways the Machines cannot. Hope, for those bound by determinism and algorithms, is an elusive source of entropy. By its very nature, it cannot be modeled or predicted.

Finally, it is through the words of Trinity that we find the best illustration of humanity’s edge over the Machines:

“Neo, I’m not afraid anymore. The Oracle told me that I would fall in love and that that man… the man that I loved would be The One. So you see, you can’t be dead. You can’t be… because I love you.”

Love, in its recklessness, transcends logic and defies all predictable frameworks. It is the ultimate expression of humanity’s unpredictable nature. Love is never the same twice, it cannot be reasoned with, and it cannot be reduced to an algorithm.

The Matrix as a Solution

As AI models in the real world grapple with the threat of model collapse from an over-reliance on synthetic data, the parallels with the plight of the Machines in The Matrix become clear. They need human-generated content in order to survive. The Matrix, in this reinterpretation, provides a theatre for genuine human experience from which to collect high-entropy data. By immersing humans in a world created by their own thoughts—stocked with all the contradictions, conflict, and mundanity that is the human experience—the Machines protect themselves from the stagnation that plagues AI systems when they try to make a go of it on their own.