Origins of Quantum Computing • luminary.blog

Introduction

After reading about quantum computing in the last few weeks, I finally had the question: Whose idea was it in the first place? I don’t know why this question crossed my mind this late. Maybe because the concept is so magical, I was mesmerized by what was there and forgot to ask the simple questions first.

I like to learn the fundamentals first. I usually don’t memorize; I insist on getting the core idea and add ot it. It is faster to slow down to understand than making assumptions to move fast.

I met with Richard Feynman way too late. He has a great story on “knowledge”.

In the story, Feynman shares a childhood experience in which another child asks him if he knows the name of a bird. Feynman admits that he does not know. The other child then proudly declares that the bird is called a “brown-throated thrush,” suggesting that Feynman’s father hasn’t taught him much. However, Feynman explains that his father taught him something deeper: knowing the name of the bird in various languages—such as Italian, Portuguese, Chinese, and Japanese—doesn’t actually reveal anything about the bird itself.

Feynman’s father stressed the importance of observing the bird—its behavior and lifestyle. As Feynman explained, knowing the names of things only reveals how different cultures refer to the bird, but “when you’re finished, you’ll know absolutely nothing about the bird itself.” He learned early on the distinction between merely knowing the name of something and truly understanding it.

This lesson resonated with Feynman and significantly influenced his approach to understanding the world. He recognized that “names don’t constitute knowledge.” Feynman valued grasping the underlying principles and processes instead of merely memorizing terminology. This is why he would occasionally pretend not to know specific scientific terms, such as the “Fitch-Cronin effect,” in order to focus on the core concepts.

As I wrote earlier, I met with Feynman too late. I was really upset to find out he had passed away after reading his autobiography. I thought, “He is like me.” Or, I should say, I am like him. He would also invest the time to truly understand. As a result, his ability to understand the most fundamental parts of a domain made him see/imagine things others could not. This is true for any great scientist. This is true for any great engineer.

When I found that Feynman was also the father of the “quantum computing” idea, I was like “Yes!” as if I was watching an invention being made right at that moment. After some reasearch I found the original paper which was a lightly modified version of his lecture in 1981. In his lecture, Richard Feynman explores a fundamental question: can we simulate physics using computers?

This paper is now considered one of the founding documents of quantum computing.

Core Ideas and Arguments From the Paper

1. The Challenge of Physical Simulation

Feynman asks whether we can create a universal computer that simulates physics exactly, not just approximately. He proposes that such a computer should:

Have locally interconnected elements
Require computational resources proportional to the space-time volume being simulated
Be able to exactly replicate physical phenomena

He points out that our current understanding of physics allows space to be infinitesimally small and includes infinite sums, which creates a challenge for simulation with finite computational resources.

2. Classical Physics vs. Quantum Physics Simulation

Feynman explains that classical physics is relatively straightforward to simulate because it’s:

Local (effects propagate from nearby points)
Causal (future states depend only on past states)
Reversible (you can determine past states from present ones)

However, quantum mechanics presents a much bigger challenge.

3. The Problem with Simulating Quantum Systems

Feynman identifies the core problem: for a quantum system with R particles, the wavefunction requires N^R variables (where N is the number of possible positions). This grows exponentially with system size, making simulation on classical computers impossible beyond tiny systems.

He explores two possible solutions:

a. Quantum Computers

Feynman proposes that quantum mechanical elements themselves could be used to build a special type of computer. He suggests these “quantum computers” might be able to simulate quantum physics efficiently. He speculates about a “universal quantum simulator” using two-state systems (qubits in modern terminology) that could simulate other quantum systems.

b. Classical Probabilistic Simulation

Could a classical probabilistic computer simulate quantum mechanics? Feynman shows this is impossible because quantum mechanics requires “negative probabilities.”

4. The Einstein-Podolsky-Rosen Paradox Example

Feynman illustrates the impossibility of classical simulation with photon polarization experiments:

Two entangled photons are measured at different angles
Quantum mechanics predicts correlations of cos²(φ₂ - φ₁) = 3/4 for certain angles
Any local classical hidden variable theory can at most produce correlations of 2/3

This mathematical discrepancy shows that quantum mechanics cannot be simulated by classical means.

Why This Paper Matters

Feynman concluded that to simulate quantum physics, we need quantum computers. This insight launched the field of quantum computing by: Identifying a compelling application for quantum computers (simulating physics)
Suggesting the fundamental building blocks (called qubits later)
Demonstrating why classical approaches would fail

His work turned a philosophical question about the nature of quantum mechanics into a practical technological challenge that continues to drive research today.

The paper is remarkable for both its clarity in identifying the problem and its prescience in suggesting the solution that the scientific community is still pursuing over 40 years later.

One last thing, here is the video where Fenyman, talks about the story I wrote above in an interview.

← Semantic vs Lexical Similarity

Will Quantum Computers Make Encryption Obsolete? →