I noticed that Vitalik Buterin recently published an intriguing technical proposal about the future of Ethereum, outlining two radical changes that could reshape the network's core architecture.

The first change concerns the state tree. Buterin suggests moving from the current Merkle Patricia tree to a more efficient binary structure, which is already in development in EIP-7864 by a team of developers. The idea is simple but powerful—reducing Merkle branches by four times means lowering verification costs and saving bandwidth proportionally.

If we use blake3 as the hashing function, we get an additional 3x improvement. Poseidon could be up to 100 times faster, but Buterin warned that it requires more security audits before practical deployment.

The other exciting part is the new storage system. The binary design combines 64 and 256 consecutive slots, allowing smart contracts to read data more efficiently. The result? Saving over 10,000 gas per transaction for high-traffic decentralized applications—this includes most active contracts today.

The second change is more ambitious. Buterin wants to replace the EVM with RISC-V, the architecture already used in ZK proofs. The logic is clear—if the tools are already available in RISC-V, why not build directly on it instead of using complex assemblies? A RISC-V interpreter requires only a few hundred lines of code compared to the increasing complexity of the EVM.

The plan progresses in three phases: first, limited use of precompilations; then allowing new contracts; and finally, retiring the EVM and re-implementing it as a smart contract. Backward compatibility is fully preserved.

Interestingly, these two changes together account for about 80% of the hurdles in efficiently generating ZK proofs. This means the future of Ethereum could be entirely different from what we imagine now, especially regarding scalability and efficiency.