Optimism co-founder Karl Floersch says AI agents could unlock mainstream Web3 adoption

Most people have never touched a smart contract. The interfaces are confusing, the stakes are high if you get something wrong, and the mental overhead of managing crypto wallets while trying to actually do something productive is, generously speaking, a lot. Karl Floersch thinks AI agents are the answer to that problem.

Floersch, co-founder of Optimism and one of the more influential architects of Ethereum’s Layer-2 ecosystem, laid out a vision in a recent Cointelegraph video where autonomous AI systems handle the heavy lifting of Web3 interaction. Trading, smart contract deployment, economic coordination across Ethereum’s network, all of it managed by agents operating on behalf of users who would rather not learn Solidity to participate in the decentralized economy.

What Floersch is actually arguing

Floersch’s argument goes further than just user experience. He positions AI agents as potentially the first class of participants to fully utilize the cryptoeconomic mechanisms that smart contract platforms were designed for. Coordination tools, reputation systems, incentive structures built into protocols, these were always theoretically powerful. The problem was that humans are slow, error-prone, and disinclined to manage the constant micro-decisions those systems require. Agents are not.

Floersch has previously discussed reputation systems designed specifically for AI agents operating within blockchain environments, suggesting this is less of a passing observation and more of an ongoing research focus.

Optimism’s positioning in the agentic future

Optimism runs on the OP Stack, a modular framework that powers a growing number of enterprise and developer-focused blockchain deployments. The launch of OP Enterprise in 2026 extended that infrastructure toward institutional use cases.

If AI agents become meaningful participants in on-chain activity, Layer-2 networks like Optimism stand to benefit in a fairly direct way. More agent-driven transactions mean more throughput demand. More autonomous economic coordination means more use of the smart contract infrastructure those networks provide. Floersch’s thesis, intentionally or not, maps fairly cleanly onto a world where Optimism’s core product becomes more valuable.

What this means for the market and investors

There is also a governance angle. Optimism uses tokenized governance models, and if AI agents become meaningful stakeholders in on-chain economic activity, the question of how those agents participate in governance, or whether they should, becomes a live issue rather than a theoretical one.

The risk, as with most convergence narratives, is that the timeline is longer and the path more complicated than the optimistic version suggests. Autonomous agents operating with real economic stakes introduce new failure modes, new vectors for exploitation, and new regulatory questions that have not been answered anywhere in the world yet. An agent that makes a mistake in a smart contract does not get to call customer support.