In their February 2026 annual letter, Stripe co-founders Patrick and John Collison addressed the intersection of AI and cryptocurrency, specifically “agentic commerce” — the ability for AI agents to independently discover, decide and transact. While optimistic about crypto’s long-term role, the report highlights critical areas where current blockchain infrastructure is “not ready” for autonomous agents. The founders characterize the current state as similar to the mid-90s internet, viewing limitations not as “deal breakers” but as “engineering challenges” that must be solved before moving to Level 5 (fully autonomous) agentic commerce.
Stripe's Big Bet and Performance Bottlenecks
Stripe has acquired stablecoin platform Bridge and launched the Agentic Commerce Suite to help businesses prepare for this shift. The report identifies two primary technical bottlenecks: cost predictability and transaction throughput. From Stripe’s perspective, blockchains need to support 1 million to 1 billion TPS to progress. Ben Goertzel, CEO of the ASI Alliance and SingularityNET, views this forecast as “entirely plausible.” He notes that standard digital financial transactions already reach millions during peak hours, even when generated by humans through intermediaries. Agentic commerce changes the scale by orders of magnitude: “Instead of a single person initiating an action, we have an entire team of agents operating autonomously… an entire squad generating transactions.”
Overcoming the Blockchain Trilemma
Achieving Stripe’s envisioned scale requires overcoming fundamental hurdles beyond raw speed: balancing decentralization, scalability and security (the blockchain trilemma), enabling agents not confined to a single network, managing vast information volumes from autonomous squads, and enabling peer-to-peer settlements with “elevated decentralized identity.” Goertzel suggests the solution is not a single monolithic network, but a system of specialized networks — like a modern highway with dedicated lanes for buses, express traffic and freight. “By separating traffic, we avoid gridlock. That’s exactly the scalable architecture we need for agentic commerce: a network of shards, where each piece does one thing well and interacts seamlessly with the rest.”
Meanwhile, the high-stakes race for AI supremacy is triggering massive consolidation of power. Unlike blockchain’s decentralized nature, the AI sector is becoming an oligarchy dominated by tech titans investing billions in proprietary infrastructure. This concentration has sparked scrutiny over whether corporate gatekeeping will supersede the public good. Nevertheless, a resilient ecosystem of startups is launching a tactical counter-offensive by leveraging agility, niche specialization and open-source collaboration. These smaller entities bet that architectural diversity and ethical transparency will disrupt the monolithic status quo.
“We don’t come close to Google or Microsoft in size, but we’re reaching a scale that allows us to compete more effectively. One of our ‘secret sauces’ is the power of diversity. Being decentralized allows us to bring together people, communities, AI algorithms, and datasets from across the globe, in contrast to the monolithic approaches taken by large centralized entities,” Goertzel said. He added that this strategic diversity becomes particularly powerful as many leading researchers realize that simply building larger LLMs will not lead to AGI. “This is something we understood from the very beginning, which has guided our Hyperon approach to AGI and superintelligence.”
To demonstrate its belief in the “cosmopolitan approach,” SingularityNET, alongside AGI Society, has organized this year’s AGI-26 conference to explore diverse interpretations of general intelligence. Goertzel also shared thoughts on minimizing the risk of AI being controlled by a few entities or governments: “We need open, decentralized, and democratic methodologies across the entire AI pipeline: deploying and running AI systems at scale, ensuring data is supplied fairly, teaching AI systems broad human values, and making collective decisions about their development.” According to Goertzel, combining open-source code with decentralized infrastructure and governance ensures that AI remains transparent, widely accessible, and “highly beneficial for humanity and other sentient beings.”