The claims
Google's Sycamore processor in 2019 performed random circuit sampling — a task with no practical use — in about 200 seconds, against an estimated 10,000 years on the best classical supercomputer. The company called it quantum supremacy; others preferred quantum advantage.
Photonic experiments in China followed with boson sampling. Each claim came with a classical baseline that was itself an estimate, and each baseline was attacked and improved.
Why the baselines keep moving
Random circuit sampling has structure that classical tensor-network algorithms exploit. Improved simulations in 2024 reproduced Google's original benchmark on ordinary hardware in minutes, and later quantum runs were matched or beaten again.
This is not cheating; it is how the comparison works. 'Quantum advantage' is a statement about the best known classical algorithm at a moment in time, and classical algorithm research is cheap and fast.
Not useful yet: sampling benchmarks have no application. Advantage on a task someone wants done — factoring, chemistry, optimisation — has not been demonstrated.
Error correction is the gate. The meaningful milestone is logical qubits that are more reliable than the physical ones beneath them, sustained over many cycles. Several groups reported progress on this in 2024–25, which is the most important development in the field.
The honest framing: quantum computers will likely be useful for a narrow set of problems — quantum chemistry, materials, some cryptanalysis — and not for most of what computers do. Announcements that say otherwise are marketing.
The archive labels this live debate: the physics is real, the engineering is advancing, and the word 'advantage' is doing more work than the evidence.
Why it is in the archive
Because 'quantum computers will break all encryption next year' and 'quantum computing is a scam' are both wrong, and the truth is specific and interesting.
Related: time crystals, monopoles in spin ice — quantum effects made tangible.
Sources for this page
- Wikipedia, “Quantum supremacy” — claims and classical responsesen.wikipedia.org
- Arute et al. (2019), Nature — quantum supremacy using a programmable processordoi.org
- Google Quantum AI, error-correction resultsquantumai.google
- Wikipedia, “Quantum error correction” — the threshold problemen.wikipedia.org