A paper claiming to crack RSA-2048 encryption with just 5,000 quantum bits went viral last week. Security researchers panicked. Cryptographers started sweating. And then Scott Aaronson actually read the paper.
The so-called "JVG algorithm" doesn't work. Not in a "needs refinement" way or a "promising but early" way. It doesn't work in a "violates basic computational principles" way. Aaronson's breakdown is worth reading in full, but here's the core flaw: the algorithm precomputes exponentially many values using classical computing, then uses quantum computing to look them up.
Think of it like this. Imagine claiming you've invented a significant way to find a needle in a haystack. Your method: carefully examine every piece of hay one by one using traditional methods, write down the location of the needle, then use your significant technique to read that location off a piece of paper. Technically you found the needle. But the "significant technique" did essentially none of the work.
Why This Matters Beyond One Bad Paper
The reason this paper spread wasn't because cryptographers vetted it and sounded the alarm. It spread because it sounded plausible to people outside the field. "5,000 qubits" is a number that feels achievable - current quantum computers are approaching that scale. "RSA-2048" is a specific, widely-used encryption standard. The claim had just enough technical detail to sound credible without being easy to verify.
Aaronson's post isn't just a debunking - it's a reminder that quantum computing claims need the same sceptical scrutiny as any other breakthrough announcement. The field is genuinely advancing, which makes it easier for bad claims to hide among the real progress.
The Exponential Problem
Here's the technical issue in slightly more detail. Factoring large numbers is hard because the search space grows exponentially. To crack RSA-2048, you need to search through roughly 2^2048 possible factors. That's a number so large it's meaningless to try to visualise.
Quantum computers are theoretically useful here because algorithms like Shor's algorithm can factor numbers in polynomial time - vastly faster than classical approaches. But they still need to do quantum work proportional to the problem size. You can't just precompute all the answers classically and claim the quantum computer solved it.
The JVG paper does exactly that. It offloads the exponentially hard work to classical precomputation, which means it's not solving the hard problem at all. It's just moving it somewhere less visible in the paper's description.
What Actually Works
Real quantum factoring research is advancing, but the numbers are nowhere near RSA-2048 yet. Current demonstrations factor small numbers - think double digits, not thousand-bit encryption keys. The gap between "factoring 21" and "factoring RSA-2048" is not incremental. It's exponential.
Shor's algorithm is mathematically sound. The challenge is building quantum computers stable and large enough to run it at scale. That requires advances in error correction, qubit coherence, and fault-tolerant quantum gates. Progress is real, but it's measured in years or decades, not months.
For Anyone Building With Encryption
If you're responsible for security infrastructure, here's what this means practically. RSA-2048 is not broken. The paper claiming to break it is fundamentally flawed. But the reason it went viral is that people are genuinely uncertain about quantum timelines, and that uncertainty creates space for bad claims to spread.
The sensible approach remains what it's been: monitor quantum progress, plan migration paths to post-quantum cryptography, but don't panic over individual papers without rigorous peer review. When a real breakthrough happens, it won't be announced in a paper with basic mathematical errors. It'll be verified, replicated, and thoroughly scrutinised before it's accepted.
Aaronson's debunking is a service to the field. It's also a reminder that hype and fear spread faster than rigorous analysis. If a claim sounds too dramatic to be true, read the paper. Or find someone like Aaronson who already has.