Foresight News reports that Cardano founder Charles Hoskinson tweeted that Vitalik Buterin, having invested too much in hash-based cryptography research, is trying to convince the industry that lattice-based cryptography is unreliable. The arguments against lattice-based cryptography are merely an analogy to the general number field sieve (GNFS), intuition about "structure," and a key amplification factor lacking formulaic support, which has not been validated over the past decades.
He pointed out that Vitalik's post contains an error in the way GNFS complexity is written; LLL, BKZ, and sieving have long been incorporated into the security parameter evaluation of ML-KEM and ML-DSA, and quantum attacks on ideal lattices have not affected these two modular lattice schemes. Hash functions also have mathematical structures that can be studied and attacked; for example, Poseidon uses low-degree polynomial mappings. Amplifying keys by 10 times has no complexity formula to support it: constant improvements correspond to proportional adjustments in dimension, not a fixed multiplication by 10. He also said that against the backdrop of growing concern over "harvest now, decrypt later" threats, excessively questioning already deployed hybrid ML-KEM schemes may slow the migration to post-quantum cryptography, leaving more communications data continuing to face the risk of future decryption by quantum computers. At the same time, he acknowledged that hash-based signature schemes have application value, but believes they are difficult to substitute for the common functions of elliptic curve cryptography.