Quantum Computing
The Day a Computer Did Something No Human Could Ever Check
In 2019, Google claimed its quantum chip had solved in 200 seconds a problem that would take the world's fastest supercomputer 10,000 years — and the most unsettling part wasn't the speed, it was that almost no one could verify it.
The Idea
Quantum supremacy — now often called quantum advantage — is the moment a quantum computer performs a specific task that no classical machine can do in any practical timeframe. The word 'supremacy' sounds like a general conquest, but it's more like a narrow, carefully staged duel. You pick a problem, run it on the quantum machine, then ask: could anything built from ordinary transistors match this? When the answer is genuinely no, you've crossed a threshold. What makes quantum machines different isn't just speed. Classical computers process bits — each one either a 0 or a 1. Quantum computers use qubits, which exploit two principles from quantum mechanics: superposition (a qubit can be in a blend of 0 and 1 simultaneously) and entanglement (qubits can be correlated in ways that have no classical equivalent). This lets a quantum processor explore vast numbers of possible states in parallel, rather than testing them one at a time. The catch — and it's a significant one — is that achieving supremacy on a contrived benchmark doesn't mean quantum computers are ready to crack encryption, simulate drug molecules, or optimise global supply chains. Those applications require not just raw quantum power but also error correction, stability, and scale that current machines don't yet have. Supremacy is a proof of concept, not a product. Think of it less like the Wright Brothers building a commercial airline, and more like them lifting off the ground at Kitty Hawk — unmistakably real, but still very far from a flight to Tokyo.
In the World
On 23 October 2019, Google published a paper in Nature announcing that its 53-qubit processor, called Sycamore, had completed a random circuit sampling task in 200 seconds. Their estimate: the same calculation would take Summit — IBM's supercomputer, then the most powerful classical machine on earth — roughly 10,000 years. The claim landed like a thunderclap, but the counterpunch came almost immediately. IBM, with obvious competitive skin in the game but also legitimate technical grounds, argued that Google had underestimated Summit. With better algorithms and more disk storage, IBM said, Summit could complete the same task in around two and a half days — not 10,000 years. Still slower than Sycamore, but hardly an epochal gap. This exchange revealed something important: quantum supremacy isn't a clean finish line. It's a moving target, because classical computers can always be given better algorithms. In 2022, a team at the Chinese Academy of Sciences published a classical algorithm that could simulate Google's circuit sampling task far more efficiently than anyone had assumed possible. The goalposts shifted again. Google, for its part, has continued pushing. In late 2024, it unveiled a new chip called Willow, claiming it could solve a benchmark problem in under five minutes that would take classical supercomputers an incomprehensibly long time — longer, they noted, than the age of the observable universe. Whether that framing survives the next round of classical counter-algorithms remains an open question, and that's precisely the point.
Why It Matters
You don't need to understand qubits to have a stake in where this is heading. The reason quantum computing attracts serious attention — from governments, defence agencies, and major technology firms — is that some of the problems it could eventually solve are foundational to how digital life works. The encryption protecting your messages, financial transactions, and medical records relies on the assumption that certain mathematical problems are computationally hard. A sufficiently powerful quantum computer, running an algorithm called Shor's algorithm, could factor large numbers exponentially faster than any classical machine — which would render much of today's public-key cryptography obsolete. We're not there yet, not remotely. But the timeline is uncertain enough that intelligence agencies are already worried about 'harvest now, decrypt later' attacks — adversaries collecting encrypted data today, banking on being able to break it once quantum hardware matures. Governments are already mandating transitions to post-quantum cryptographic standards. Knowing what quantum supremacy actually means — a narrow, contested, benchmark-specific milestone, not a general takeover — gives you a more accurate map. When you see the next headline declaring that a quantum computer has broken the internet, you'll know to ask the right questions: what specific task, what comparison, and what did the critics say?
A Question to Ponder
If a computation's result is too complex for any classical machine to verify, how do we decide whether to trust it — and what does that say about how we've always decided to trust the machines we build?
Get a new one of these every morning.
Start learning with Thinkable