The Verification Problem Nobody Talks About
Quantum computing has a credibility problem that has nothing to do with the hardware itself. For years, researchers have mathematically proven that certain algorithms run exponentially faster on quantum machines than on classical ones – but today’s quantum computers either cannot execute those algorithms at all, or can only run stripped-down versions that a regular computer handles just as well. That gap between theoretical promise and practical demonstration has quietly become the field’s most stubborn obstacle.
The deeper issue is almost paradoxical: if a quantum computer produces a result that no classical machine could generate in any reasonable timeframe, how do you confirm the answer is actually correct? You can’t check it against a classical benchmark that doesn’t exist, and today’s quantum hardware is error-prone enough that a wrong answer is a genuine risk, not a theoretical one.
IBM is trying to close that credibility gap.
On Thursday, IBM announced three new entries to its quantum advantage tracker – a framework the company launched specifically to document cases where quantum hardware demonstrably outperforms classical computing, with results that can be trusted. Each of the three new entries takes a different approach to the twin problems of error management and result validation. Together, they represent IBM’s most direct public argument yet that quantum advantage isn’t just a future promise.
Why “Quantum Advantage” Needs a Tracker in the First Place
The phrase “quantum advantage” gets used loosely in this industry. Companies have claimed it before in ways that didn’t survive scrutiny – either because a better classical algorithm eventually matched the quantum result, or because the benchmark problem was too narrow to mean anything outside a lab. IBM’s tracker is a direct response to that history. By publishing a running log of validated cases, the company is betting that transparency will hold up better than one-off announcements.
IBM’s Jay Gambetta put the stakes plainly: “Trusted computing when you can do classical simulations is irrelevant. Trusted computing when you can’t do classical simulations is a big deal.” That distinction matters more than it might first appear. Plenty of quantum demonstrations over the past decade have operated in the regime where classical simulation is still possible – which means they were never truly out of reach, and any error in the quantum result could at least be caught. Once you move past that boundary, the error question becomes genuinely hard. If you can’t simulate the result classically, you need a different method entirely to confirm the quantum machine got it right.
The three new tracker entries each use a distinct method to address this. IBM hasn’t published full technical details in the announcement itself, but the company’s framing makes clear that the validation strategy – not just the speed of computation – is what distinguishes these cases from earlier quantum advantage claims. The fact that there are three separate approaches suggests IBM is deliberately testing different paths to trusted quantum output rather than committing to a single methodology. That’s a notable strategy for a company that also sells quantum hardware commercially.
Noise, Errors, and the Hardware Reality
Current quantum computers are what researchers call “noisy” – meaning errors accumulate during computation in ways that don’t happen with classical transistors operating at room temperature under normal conditions. Quantum bits, or qubits, are fragile. They lose their quantum state through a process called decoherence, and the longer a computation runs, the more errors pile up. This isn’t a flaw unique to IBM; it’s a property of quantum hardware at this stage of development across every major player in the space, including Google, Microsoft, and a growing list of startups.
Error correction is theoretically solved – there are well-established quantum error correction codes that can protect computation – but implementing them at scale requires vastly more physical qubits than today’s machines contain. The workaround IBM and others have pursued is “error mitigation,” a set of techniques that reduce the impact of noise without full error correction. It’s an imperfect fix, but it has allowed researchers to push results into territory that gets closer to genuine quantum advantage without waiting for hardware that’s still years away. The three new tracker entries presumably rely on some combination of these approaches, each tuned to a different type of calculation.
What makes IBM’s announcement worth watching isn’t that quantum computers are suddenly error-free – they aren’t. It’s that the company is building a public accountability structure around the claims it makes, at a moment when the broader industry still lacks any neutral third-party standard for what “quantum advantage” actually requires. Whether IBM’s own tracker qualifies as rigorous enough is a fair question, and one that researchers outside the company will inevitably pressure-test.
There’s a version of this story where IBM’s tracker becomes the industry benchmark – the standard other companies reference when making their own claims, the way NIST standards shape cryptography or the way Wi-Fi certification shaped wireless hardware. There’s another version where competitors publish rival trackers with different methodologies, and the field ends up more fragmented than before. What IBM has placed on the table Thursday is not a solved problem but a framework for what solving it might look like – and the difference between those two things is exactly what Jay Gambetta’s comment about “trusted computing” was pointing at. When you can’t verify the answer any other way, the process you used to get there is the only thing you have left.
