Google Quantum AI has shown on its Willow chip that one of the costliest steps in quantum computing can be done with roughly 40 times fewer errors. Wallets are not at risk today, but results like this bring closer the day when protecting Bitcoin from quantum computers stops being a theoretical question.
What exactly did Google show?
In a paper published in Science on Thursday, the team described an experiment with a technique called magic state cultivation. On Willow, it cut errors in so-called magic states by about 40 times. The logical fidelity of those states reached a record 99.99%.
A magic state is a special resource a quantum computer needs for its harder operations. Think of a single-use fuel cell: each one fires only once. Qubits themselves (quantum bits that hold a blend of 0 and 1) are fragile, since noise or heat wrecks a calculation. That is why error correction spreads one unit of information across many qubits.
How does it work, and why was it expensive before?
The standard method was laid out by Sergei Bravyi and Alexei Kitaev back in 2004. It is called distillation: many noisy states are boiled down into a few clean ones. It works, but it needs dedicated qubit "factories" that do nothing else. Google estimates clean states can eat up to 95% of a machine's capacity.
Cultivation, or "growing" the states, acts as a filter. Here is how it looks.
- Start: a rough, noisy state is injected into a small patch of the chip.
- It then runs through a chain of checks.
- Any attempt where an error is found gets thrown out.
- The survivors are grown to the required quality.
The price is the number of attempts. Only about 8% pass every check, so the chip discards more than nine in ten. But no qubit factory is needed, and that is what makes a future machine cheaper.
Does this threaten Bitcoin wallets right now?
No. The experiment used at most 54 of Willow's 105 qubits. Breaking an RSA-2048 key would take magic states with an error of about one in 10 billion, roughly a million times cleaner than today's result.
The trend is still worrying. Bitcoin relies on elliptic curve cryptography, which stops anyone from deriving a private key from public data. In March, Google teamed up with Ethereum Foundation researcher Justin Drake and Stanford cryptographer Dan Boneh to put the cost of breaking it at under 500,000 physical qubits. The team names running cultivation at a larger scale as its next step.
How much bitcoin is exposed?
The same day, nine crypto companies unveiled the Quantum Alliance, which is meant to develop shared security standards for Bitcoin, Ethereum and other networks. The alliance estimates that 6.4 million BTC, or 30.2% of supply, already have exposed public keys. About 1.1 million of those are attributed to Satoshi Nakamoto. Those coins cannot be stolen today, because the required computer does not exist yet.
When it will, experts disagree. Richard Carback of Postquant Labs thinks 2028 is realistic. Bryn Bennett of The Vault says the date matters less than another question: whether custodians can migrate in time if the threat arrives early.
What can be done about it?
Binance founder Changpeng Zhao and developer Jameson Lopp have both floated freezing vulnerable coins. Critics reply that a freeze breaks the principle that nobody can confiscate someone else's bitcoin. Bennett says it is a decision for the network, not the industry.
The technical fix is a single one, namely replacing the signatures that authorize transactions. New signatures may take more space on the blockchain, and the main cost lies in coordinating wallets, exchanges and developers. Starknet has already opened a seven-day vote on an independent L1 over quantum risks.
Separately, there is a scenario Justin Drake, as relayed by Decrypt, urged people to factor in now. He tells large holders to move funds to fresh addresses that have never signed a transaction, because a signature exposes the public key. Drake himself thinks AI-driven mathematics will break elliptic curves first, on ordinary hardware, in "months not years".
The industry is already moving. Coinbase argued this week that migration has to start now. Sui is testing quantum-safe transactions on a test network, and Ethereum, Zcash and Near all have transition plans.
What should an ordinary holder do?
Nothing urgent. Still, it is worth asking your wallet or exchange about its plans for moving to new signatures, rather than leaving that talk until the date is known.
Comments
Your email address will not be published. Required fields are marked *