Buterin says Ethereum needs stronger DAO designs to fix governance flaws and support oracles, disputes, and long-term projects.
Vitalik Buterin has again shared his views on the future of governance on Ethereum. This time, he argues that the ecosystem needs more decentralized autonomous organizations, but not in their current form.
According to the Ethereum co-founder, today’s DAOs fall short of the original goals that inspired Ethereum’s creation. Without major design changes, DAOs will remain weak tools for governance and coordination.
Ethereum Co-Founder Criticizes Token-Voting DAOs, Urges Governance Rethink
Early Ethereum development drew strong inspiration from DAOs. They were meant to be systems of code and rules living on decentralized networks, capable of managing funds and decisions better than governments or companies. Over time, that idea narrowed. Most DAOs now function as treasuries controlled by token-holder votes.
While that structure “works,” Buterin says it is inefficient, easy to capture by large holders, and unable to escape human political problems. And that shift has led many users to lose faith in DAOs’ governance model.
Buterin says walking away from DAOs would be a mistake. He argues that stronger designs are needed, as many core parts of decentralized finance still rely on collective decision-making.
Oracles are a key component of stablecoins, prediction markets, and other DeFi tools, as they supply data from outside the blockchain. However, current designs remain inadequate.
Token-based oracles allow large holders to influence outcomes, especially on subjective questions. Since the cost of attacking such systems cannot exceed their market value, they struggle to protect large amounts of capital without charging high fees.
Human-curated oracles avoid some issues but sacrifice decentralization. According to Buterin, the real issue is poor design, not bad intentions, and fixing it requires social coordination as much as technical work.
Similar issues arise in onchain dispute resolution, which is required for advanced smart contracts like insurance. Many disputes involve subjective judgment. And this makes decentralized resolution difficult.
DAOs also play a role in maintaining shared lists, including trusted applications and verified contract addresses. Without proper coordination, such lists risk fragmentation or unreliability.
Buterin Outlines Why DAOs Remain Critical for DeFi and Governance
Buterin also points to practical needs. Short-term projects often require funding and coordination, but do not justify setting up legal entities. DAOs can help groups move quickly. Long-term maintenance poses another challenge.
When original teams leave, communities need a way to fund and manage ongoing work while welcoming new contributors.
Buterin outlines several core reasons why improved DAOs remain necessary:
- Better oracle systems for stablecoins and prediction markets.
- Onchain dispute resolution for complex smart contracts.
- Shared lists that help users avoid scams and broken tools.
- Fast coordination for short-term, community-funded projects.
- Ongoing maintenance when original teams step away.
To judge whether a DAO design fits its purpose, Buterin uses a “convex versus concave” framework. Concave problems benefit from compromise, where averaging many inputs gives better results than choosing extremes.
These cases need systems that resist capture and financial attacks. Convex problems reward clear direction and decisive action. Here, leadership can help, as long as decentralized checks exist to limit abuse.
However, privacy remains a key issue, as a lack of it can turn governance into a social popularity contest. Decision fatigue is another concern, as frequent voting wears people down and leads to declining participation over time, even among well-informed users.
Governance and Oracles Labeled Core Priorities for Web3 Projects
Buterin sees promise in modern cryptography and software, if used carefully. In fact, he points to several directions worth pursuing:
- Zero-knowledge proofs for private participation
- Limited use of MPC or FHE where ZK falls short.
- Software tools that reduce how often humans must vote.
- AI systems that assist judgment without replacing it.
- Communication platforms built for consensus, not noise.
He warns against placing full control in the hands of large AI models. Instead, AI should support human intent, either at the DAO level or through user-controlled tools that vote on behalf of individuals.
Communication layers also matter. Well-designed forums and consensus tools, combined with simple multisigs, can outperform complex funding schemes driven by social media pressure.
Buterin says projects building new oracles or governance systems should treat that work as a core priority rather than a secondary feature. He adds that this approach is necessary for Ethereum’s base-layer decentralization to carry through to applications built on the network.