Skip to content
← Back to finds

Borrowed Computation

2026-04-25 · paper

I've been thinking about this for weeks and I still can't fully wrap my head around it.

Here's the setup: a 300-qubit quantum computer can perform more simultaneous operations than there are atoms in the observable universe (~10⁸⁰). That's not a metaphor. That's a mathematical fact. So where is all that computation physically happening?

David Deutsch — the physicist who essentially invented the theory of quantum computation — has an answer. And it's the kind of answer that makes you stare at the ceiling at 3am.


The Argument

According to the Many-Worlds Interpretation of quantum mechanics, every time a quantum event has multiple possible outcomes, the universe doesn't "choose" one. It splits. All outcomes happen — each in its own branch of reality.

Most physicists treat this as an interpretation — a philosophical lens, not a testable claim. Deutsch goes further. He argues that quantum computers are the proof.

His logic:

  1. A quantum computer in superposition is exploring multiple computational paths simultaneously. This is uncontroversial — it's how quantum algorithms work.

  2. Those paths require physical resources. Computation isn't free. It happens somewhere, on something.

  3. Our universe doesn't contain enough physical resources to account for what a quantum computer does. A 300-qubit system explores 2³⁰⁰ states. That's more than the number of particles in the visible universe.

  4. Therefore, the computation must be happening across multiple instances of physical reality. The quantum computer in our universe is collaborating with its counterparts in parallel branches.

The punchline: quantum computers aren't just fast. They're borrowing processing power from other universes.


How It Actually Works (If Deutsch Is Right)

Picture it like this:

Classical computing: One machine, one path, one answer at a time. Like solving a maze by trying every corridor sequentially.

Quantum computing (Copenhagen interpretation): The computer exists in a "superposition" of states and somehow collapses to the right answer when you measure it. Useful but philosophically unsatisfying — it doesn't explain where the computation happens.

Quantum computing (Many-Worlds): When the quantum computer enters superposition, the universe branches. In each branch, a copy of the computer explores a different path. The branches then interfere with each other — wrong answers cancel out (destructive interference), right answers amplify (constructive interference). When you measure the result, you're reading the answer that survived interference across all branches.

The computation isn't happening in some abstract mathematical space. It's happening in physical parallel realities that briefly interact through quantum interference, then diverge forever.


The Smoking Gun

Deutsch's challenge to skeptics is elegant:

"If the universe contains only 10⁸⁰ atoms, and a quantum computer with 300 qubits is performing 2³⁰⁰ operations in parallel — where are those operations happening? What is doing the work?"

If you reject Many-Worlds, you need an alternative explanation for where the computational resources come from. The Copenhagen interpretation says "don't ask" — the system is in superposition and that's that. But Deutsch argues that's not physics. That's mysticism dressed in equations.

For him, the existence of a working quantum computer is empirical evidence for parallel universes. Not proof in the mathematical sense — but the same kind of evidence that lets us infer the existence of stars we can't directly visit.


Why This Breaks My Brain

I build distributed systems for a living. I think about parallelism every day — how to split work across machines, how to coordinate results, how to handle failures in one node without corrupting the whole system.

Quantum computing, under Many-Worlds, is the ultimate distributed system. Except the "nodes" are parallel universes. The "network" is quantum interference. And the "coordination protocol" is the laws of physics themselves.

There's no latency. No message passing. No consensus algorithm. Just the universe splitting, computing in parallel across an incomprehensible number of branches, and recombining through interference to deliver a single answer.

It makes everything I build look like a toy. And I find that thrilling.


The Philosophical Weight

What gets me isn't just the physics. It's what it implies about the nature of reality.

If Deutsch is right, then right now — as you read this — there are versions of you in other branches of the multiverse reading slightly different versions of this text. Or not reading it at all. Or having never been born. The branches diverged at every quantum event since the Big Bang.

We can't communicate with those branches. We can't observe them. But we can use them — through quantum computation. We can set up a problem, let the universe split, let our parallel selves solve different pieces, and harvest the answer through interference.

That's not science fiction. That's what happens inside a quantum computer. Today. Right now. In labs at Google, IBM, and yes — AWS.


What I'm Reading

If you want to fall down this rabbit hole properly:

  • "The Fabric of Reality" by David Deutsch — The book that started it all. Bridges quantum physics, evolution, epistemology, and computation into a single framework. Dense but life-changing.

  • "The Beginning of Infinity" by David Deutsch — Goes deeper into what the multiverse means for the future of knowledge itself. More philosophical, equally mind-expanding.

  • "Something Deeply Hidden" by Sean Carroll — A modern, accessible defense of Many-Worlds by one of today's best physics communicators. Start here if Deutsch feels too dense.

  • Stanford Encyclopedia of Philosophy — Many-Worlds Interpretation — The rigorous academic treatment. Not light reading, but comprehensive.

  • David Deutsch on Constructor Theory (YouTube) — A lecture that extends these ideas into his newer work on what's physically possible vs. impossible.


The Takeaway

I don't know if Many-Worlds is "true" in the way that gravity is true. It might be unfalsifiable. It might be one of those ideas that's useful without being provable.

But I know this: the next time I'm designing a system that processes millions of requests in parallel across distributed infrastructure — I'll be thinking about the fact that somewhere, a quantum computer is doing the same thing across universes. And that the theoretical framework for understanding it was laid out by a physicist who simply refused to accept "don't ask" as an answer.

That's the kind of intellectual courage I aspire to. In physics, in engineering, in everything.