People have been claiming for years that the cryptography underpinning bitcoin is about to break. This week, Ethereum researcher Justin Drake urged holders to start planning a move to fresh addresses because AI might find a way to recover private keys from exposed public keys within months. That is a hell of a recommendation to make without evidence that the underlying mathematical problem has become easier to solve.
Cryptographer Yehuda Lindell’s response gets to the heart of it: AI proving difficult theorems does not mean it can break elliptic-curve cryptography. These predictions prey on the fact that very few people can evaluate the math for themselves. Mine was more succinct.
I believe ECDSA will never be broken. AI is not going to conjure a classical shortcut because someone made a dramatic X post. Shor’s algorithm concerns a different threat: a cryptographically relevant quantum computer that nobody has built. I’m not moving bitcoin in response to an imagined attack. Moving valuable coins in a hurry introduces a real risk of losing them today.