Yesterday I wrote that Vitalik advises against rushing wallet migration. And I missed a number from his own text, which is now the cause of a spat.
Vitalik: if AI delivers “50 years of math in 2 years,” keys might need to be increased by as much as tenfold.
Hoskinson replied (according to Coinpedia): the 10x rule is “numerology,” and Vitalik himself is “bag-holding” hash cryptography. Against a slightly better attack, he says, keys about a quarter larger should be enough.
Funny, because last December Hoskinson himself explained: hashes are Ethereum’s bet, lattices are Cardano’s bet. So both are bag-holding—$ETH with one, $ADA with the other.
The only thing AI has actually found against lattice-based cryptography is Anthropic’s attack on HAWK. There, keys need to be doubled; NIST’s other candidates weren’t affected.
Doubled. Not tenfold, and not by a quarter.
The argument is loud, but for now the proof is still with a third party.
#PostQuantum #Cardano
Vitalik: if AI delivers “50 years of math in 2 years,” keys might need to be increased by as much as tenfold.
Hoskinson replied (according to Coinpedia): the 10x rule is “numerology,” and Vitalik himself is “bag-holding” hash cryptography. Against a slightly better attack, he says, keys about a quarter larger should be enough.
Funny, because last December Hoskinson himself explained: hashes are Ethereum’s bet, lattices are Cardano’s bet. So both are bag-holding—$ETH with one, $ADA with the other.
The only thing AI has actually found against lattice-based cryptography is Anthropic’s attack on HAWK. There, keys need to be doubled; NIST’s other candidates weren’t affected.
Doubled. Not tenfold, and not by a quarter.
The argument is loud, but for now the proof is still with a third party.
#PostQuantum #Cardano