Quantum computing has a checking problem, and it's a strange one: the whole point of the machines is to solve problems that ordinary computers can't, which means there may be no ordinary computer available to grade the answer.
That's the question at the center of a new Ars Technica piece, published in July 2026 and headlined "If a quantum computer outperforms normal ones, can you tell if it's right?"
According to Ars Technica, there are many algorithms for which it has been mathematically proven that a quantum computer can generate results that would take a classical computer an unreasonable amount of time to produce. That proof is the good news — it's the formal basis for believing quantum machines offer a genuine advantage rather than a marketing one.
But it cuts both ways. If a classical machine would need an unreasonable amount of time to compute the answer, it would also need an unreasonable amount of time to recompute the answer as a check. The usual scientific reflex — run it again somewhere else and compare — stops being practical exactly at the point where the technology starts being interesting.
That leaves an uncomfortable gap between a claim of quantum advantage and independent confirmation of it, and it's a gap that grows as the machines get better.
The available sourcing here is thin — a headline and a single framing statement — so the specific verification methods researchers are pursuing aren't detailed in these items.
Why it matters: if quantum computers are eventually used for drug design, materials science, or cryptography, someone has to be able to trust the output — and trust normally comes from double-checking, which is the one thing these machines make hard.