Quantum computers are not just faster versions of the laptop you use for email. They are a completely different way of processing information, designed to solve specific problems that are currently impossible for our most powerful machines, like predicting how a new drug interacts with a human enzyme or why a battery might catch fire.
A company called PsiQuantum is attempting to build a useful, large-scale quantum computer by using light. Instead of standard silicon chips that process information using electrical currents, their system uses particles of light, known as photons, reflecting through a series of tiny optical switches. To function, this machine must be housed in massive, cold cabinets kept at temperatures not far from absolute zero. PsiQuantum is currently building sites in Chicago and Australia, aiming to prove that this hardware can finally do work that is commercially valuable.
The fundamental limit of today's computers
To understand why we even need these computers, we have to look at how nature works. The world at the level of atoms and molecules does not behave like the world of pencils and cars. Subatomic particles exist in a hazy state of possibilities rather than having one fixed location or speed. Modern computers are built to handle fixed states—they use bits that are either a 0 or a 1. Because of this, they are terrible at simulating molecules, which operate on that same hazy, multi-state logic. Scientists today have to use approximations and guesses, which take years of trial and error.
A quantum computer uses what are called qubits. Unlike a standard bit, a qubit can represent multiple states simultaneously. By manipulating these qubits to mimic the natural behavior of particles, the computer can simulate chemistry and physics directly instead of approximating it. The challenge is that these systems are incredibly delicate. If a qubit is disturbed or observed while it is working, it collapses into a single, standard state, causing an error. PsiQuantum is betting that using light particles—photons—is the best way to keep these systems stable enough to actually reach an answer while keeping the errors in check.
If this effort succeeds, the impact would be most visible in fields that require deep scientific modeling. If we could simulate exactly how a drug binds to a protein in the human body, we could cut years of lab testing down to minutes. This isn't just about speed; it is about accessing knowledge that is currently locked away by the complexity of the physical world. We are currently at a moment where these companies are transitioning from theoretical experiments to the hard, unglamorous work of engineering. Whether PsiQuantum succeeds in building a machine that is actually useful will depend on their ability to move from prototypes to systems that can do real work without breaking down.
Liked this one? The next lands at breakfast.
Every story in tomorrow's AI news, rebuilt in plain English — five minutes, sources linked, free forever.
By joining you agree to receive Article's daily newsletter — unsubscribe in one click. Privacy