Vitalik Buterin expressed optimism about the long‑term development of cryptography. He estimates there is around a 33 % probability that SNARK, Fully Homomorphic Encryption (FHE) and Indistinguishability Obfuscation (iO) will eventually achieve overhead close to 1+ε times the native computation cost. There is roughly a 60 % chance that all three can keep total overhead within 10 times. He believes at least one of these technologies will hit single‑digit overhead before the end of this decade, most likely SNARK. He also notes that certain specialized hash functions and some large‑model inference scenarios are already approaching this performance level today.