Israeli mathematician Gil Kalai maintains that quantum computers will never be able to break cryptography, according to Eli Ben-Sasson, CEO of StarkWare, a company specialized in zero-knowledge (ZK) proofs and creator of StarkNet, a second layer (L2) network of Ethereum.
Ben-Sasson clarified that he does not subscribe to that position but considered it relevant to expose it: “Quantum computers will never break cryptography… It is not my opinion, but I will explain it because it is important to raise it.”
Kalai is a mathematician at the Hebrew University of Jerusalem, an adjunct professor at Yale University in the US, and a skeptic of globally scalable quantum computing. His argument, according to Ben-Sasson, revolves around noise: Any minor disturbance (a vibration, a change in temperature, even electromagnetic radiation from the environment) can alter the state of a qubit (the quantum computing units) and give an incorrect result.
A qubit resembles a house of cards, since any interference from the environment can “bring it down”causing it to fail and returning an incorrect result. In this framework, the technique of quantum error correction It seeks to stabilize the qubits, grouping several of them so that they “monitor” each other: if one fails, the others would allow the correct value to be reconstructed.
The problem that Kalai poses is that the quantum computer itself shakes the table: to more qubitsthe more disturbances the system itself generates.
According to the argument conveyed by Ben-Sasson, That noise would not be random but correlated with the computation itself. “The noise may not be random ‘oops, I was wrong’ noise that can be averaged out. It may be noise correlated with the computation. So, the more qubits, the more noise. A bad noise, which ruins the calculation,” wrote the CEO of StarkWare.
If Kalai’s premise is correct, error correction would be ineffective at scale, and Therefore it would be impossible for a quantum computer to break systems as RSA (used by banks), elliptic curves (ECCused in networks such as Bitcoin and Ethereum) or the schemes SNARKs (cryptographic proofs that allow a calculation to be verified without revealing the data that supports it).
Recent advances complicate the premise
Two recent experiments by the company Quantinuum reported by CriptoNoticias directly contradict Professor Kalai’s idea.
The first, published last February, showed that quantum error correction crossed the so-called “break-even”: the point at which shielding the qubits improves the result rather than degrading it, something that previous techniques did not achieve.
The second, published in March, extracted 48 logical qubits (functional qubits capable of reliable calculations) from just 98 physical ones, a 2:1 ratio. The most accepted industry standard estimated that building a logical qubit required between 100 and 1,000 physicists, so the estimate of This second study would reduce the scope for building scalable quantum hardware.
Likewise, Thomas Coratger, cryptographer at the Ethereum Foundation (EF), assured that through neutral atom processors that improve connectivity between qubits, the ratio would improve by 10:1.
Quantum computing and ecosystem estimates
Justin Drake, one of the main developers of Ethereum and co-author of the paper of Google Quantum AI, raised its estimate of crypto breakout probability by 2032 from 1% to 50%. Vitalik Buterin, co-founder of Ethereum, estimates that by 2028 a quantum computer could compromise ECDSAthe digital signature system that protects Bitcoin and Ethereum transactions.
Along similar lines, Mikhail Lukin, a Harvard professor and co-founder of the Harvard Quantum Initiative, believes that fault-tolerant quantum computers could be available “at least in some form” before the end of this decade. Companies like Google, Cloudflare and Grayscale set 2029 as a horizon to complete their post-quantum migrations.
At the opposite extreme, Adam Back, co-founder of Blockstream, places the threat to “at least a decade away”and Samson Mow, CEO of JAN3, extends it to between 10 and 20 years.
Kalai’s argument, as conveyed by Ben-Sasson, does not belong in that debate about deadlines. He does not discuss when the threat will arrive, but rather warns that the physical viability of quantum hardware will not allow this technology to constitute a real threat to current cryptographic systems.