Eman Abio, Co-Founder and Chief Product Officer of Mysten Labs, the company behind the Sui blockchain, appeared in a conversation with Michaël van de Poppe to explain why he believes crypto's foundational design assumptions are wrong and what Sui is doing to replace them.
Key Takeaways
Abio: majority of internet transactions will be governed by AI agents.
Stripe report cited: 1 billion TPS needed for agentic world.
Gas fee model compared to Uber surge pricing: unrelated users competing for the same resource.
Sui launching free payment SDK this year: agentic on-chain payments, no restrictions.
Abio’s starting point is not Sui. It is the internet that is coming. He argues that the majority of transactions conducted online will soon be proxied by AI agents rather than executed directly by humans, pointing to the ad industry where bots already handle the majority of internet traffic as evidence that this transition is already underway. A Stripe report he cites projects that the agentic world will require 1 billion transactions per second. No current blockchain operates anywhere near that figure. Abio is not claiming Sui has solved this. He is framing the size of the problem that the industry needs to build toward.
Why Gas Fees Are the Wrong Model
Abio’s Uber surge pricing analogy is the sharpest critique of current blockchain economics: when a user paying gas to send money to a friend competes in the same fee market as a trader executing meme coin transactions, the pricing mechanism treats unrelated activities as fungible, which means the cost of using the network is determined by who else is using it rather than by what you are actually doing. He frames this as a structural incompatibility between crypto’s scarcity-first design and how the real world operates. Google does not charge more per search because more websites exist. The pricing mechanism scales with infrastructure, not with demand from unrelated users.
His conclusion from this observation is direct: “The idea of wallets is nonsensical. The idea of gas is nonsensical. Moving money should be free.” The statement is deliberately provocative but the argument behind it is specific. If the infrastructure can be built to scale with demand rather than price out excess demand, the gas fee mechanism is not a feature. It is a workaround for an engineering problem that has not yet been solved at scale.
The Google Pedigree Argument
The engineering decision Abio describes, building the capacity to abundantly add resources rather than managing scarcity through pricing, is the same decision Google made when it chose to buy more servers instead of charging per search, and whether Sui can execute that model at the scale the agentic internet requires is the question his argument raises without fully answering. Abio’s claim to credibility on this point is specific: the Mysten Labs team built search at Google and core infrastructure at Facebook, institutions that never designed for thousands of users because they always designed for billions. He argues that pedigree is what crypto has been missing and what Mysten Labs brings.
AI agents will automate everything in crypto. Payments. Wallets. On-chain activity.@EmanAbio breaks down why @SuiNetwork is making payments free:
"The idea of wallets is nonsensical. The idea of gas is nonsensical. Moving money should be free."
How AI changes crypto adoption… pic.twitter.com/OzHtSPRVke
— New Era Finance Podcast (@new_era_finance) May 25, 2026
What Sui Is Actually Launching
The concrete output of this engineering philosophy, in Abio’s telling, is free payments on Sui launching this year. Any developer will be able to download a payment SDK and send agentic on-chain payments for free with no restrictions or limitations. He acknowledges that offering payments for free at scale is not a simple product decision: it requires the underlying infrastructure to add resources abundantly as demand grows, which is the engineering commitment Mysten Labs made early in the project’s development.
Citing a Stripe report projecting 1 billion transactions per second for the agentic world while no current blockchain operates above a fraction of that figure is an honest statement about the size of the unsolved problem rather than a claim that Sui has solved it, and Abio’s argument is most credible when read that way. The free payment SDK is a step toward that scale. Whether the engineering decisions made early in Sui’s development are sufficient to reach it will be visible in adoption data as the agentic internet arrives.
