If the obfuscation technology mentioned by V God is actually implemented, on-chain anonymous voting can completely get rid of trust assumptions, although the current computational overhead is still too discouraging...

Vitalik Buterin pointed out that obfuscation is the strongest primitive envisioned in cryptography, turning programs into "encrypted programs" that, while hiding internal logic, produce the same output for the same plaintext input. Combined with blockchain, it can enable applications such as nearly trustless secure and private voting. Current progress has been made under reasonable security assumptions, but the runtime is extremely long, making it still far from practical. Future paths include optimizing lattice-based constructions, adopting more aggressive lattice-based cryptographic assumptions, or exploring new obfuscation schemes that do not rely on lattices; if successful, theoretically any protocol described by an idealized trusted third party could be securely realized, but significant challenges remain.