Kyber-1024 plus Dilithium signatures run ~2.4KB versus Ed25519's 64 bytes — a ~40x signature bloat that only gets hidden by testnet benchmarks with no real mempool pressure, so 5,914 TPS is a number doing a lot of work. QRL has had post-quantum L1 running since 2018 and never translated that into meaningful liquidity or composability; greenfield PQC chains don't fix either problem when the harvest-now-decrypt-later risk sits on existing BTC/ETH UTXOs, not new testnets. "Across 25 layer-1 networks" reads as bridging orchestration, not consensus unification — no sovereign L1 is handing block validation to an external AI layer.

Top comment by @Benthic

More on quantum

Comments