Vitalik Buterin Pushes Ethereum Toward Bitcoin’s Minimalist Philosophy in 5-Year Overhaul

Ethereum’s co-founder throws down the gauntlet—complexity is out, Bitcoin’s ’less is more’ ethos is in. The blockchain that brought us smart contracts might just dumb things down (for its own good).

Why now? After a decade of feature creep—DeFi degens, NFT mania, and enough layer-2 solutions to make a lasagna—Buterin wants to trim the fat. No more ’kitchen sink’ development. The target? A leaner protocol that actually scales without requiring a PhD in cryptoeconomics to use.

The irony isn’t lost here: Ethereum’s $400B ecosystem built on endless flexibility now faces its ’Kodak moment’—adapt or risk becoming the overengineered relic Wall Street never understood (but will definitely try to ETF-ify anyway).

Vitalik pushes for a simpler Ethereum consensus layer

Vitalik said Ethereum can get there by rewriting its consensus layer. He wants to replace the current beacon chain with something called 3-slot finality, which would remove slots, epochs, committee shuffling, and other moving parts.

“You could build this in 200 lines of code,” he wrote. He said it offers strong security and removes a lot of bloat from the system. The smaller validator sets also make the fork choice rule simpler.

He said Ethereum should use STARK-based aggregation so anyone can be an aggregator without needing special trust or overpaying for bitfields. The cryptography is complicated, but it’s boxed in and doesn’t screw with the whole system. That opens the door to a simpler peer-to-peer layer.

Validator functions — like entry, exit, withdrawal, key changes, and inactivity leaks — should be rebuilt to reduce the line count and make system guarantees easier to read. Vitalik said the best part about consensus is it’s not tightly connected to the execution layer, so it can evolve without breaking contracts.

But the real mess, he said, is the Ethereum Virtual Machine.

Vitalik wants to kill the EVM and move to RISC-V

Vitalik said the EVM is stuffed with outdated crap. He called it an over-engineered 256-bit machine optimized for crypto tricks that nobody uses anymore. He admitted most of the complexity came from his own choices.

The fight to remove the SELFDESTRUCT opcode was a waste of time. So was the whole EOF debate. His solution: skip the small upgrades and just replace the EVM.

He proposed switching to RISC-V or another VM like Cairo — the same ones being used by Ethereum’s zero-knowledge proof systems. “Succinct data shows this could boost efficiency by 100x,” he said. Simpler specs would mean faster execution and fewer bugs. 

Developers would get more options. Solidity and Vyper would compile to new VMs, and devs using normal programming languages could write Ethereum contracts for the first time. Most precompiles could be scrapped, except maybe for elliptic curve ops.

Vitalik admitted that RISC-V won’t be ready tomorrow. While waiting, he wants to pass quick upgrades to the current EVM, such as raising contract size limits and adding new DUP/SWAP opcodes. But the real goal is to replace the entire thing. His plan comes in four steps.

First, every new precompile must come with an onchain RISC-V version. Second, developers should be allowed to write contracts in both RISC-V and EVM. Third, Ethereum would do a hard fork to remove precompiles and replace them with RISC-V contracts. Fourth, the EVM would get re-implemented inside RISC-V and run onchain as a contract interpreter.

Vitalik said Ethereum also wastes time using different tools for the same job in different places. He pointed to erasure codes, which are needed for data availability, P2P broadcasts, and history storage. All three should use the same code. The same goes for serialization formats — Ethereum should move fully to SSZ, which is already in use in the consensus layer and works inside smart contracts.

He said Ethereum’s Merkle tree also sucks. The current hexary structure makes block proving hard. Switching to a binary tree with a better hash would make proving faster and cheaper. That same binary tree should be used for both execution and consensus.

Cryptopolitan Academy: Coming Soon - A New Way to Earn Passive Income with DeFi in 2025. Learn More