XPRIZE Founder Peter Diamandis: The Agency Economy Depends on Stablecoins and Infrastructure
XPRIZE founder Peter Diamandis discussed the infrastructure and attitude differences shaping the agency economy with Emad Mostaque, Cathie Wood, and Chandhok at Moonshots Live.
The discussion covered the foundations that AI relies on: energy, data centers, stablecoins, and faster financial infrastructure. It also addressed the disparity in attitudes toward AI between the U.S. and China, why healthcare adoption may lag behind other industries, and the complexity of humanoid robots being greater than most assumptions. The live event was hosted by the Abundance/Moonshots system, with sponsors including Circle, ARK Invest, and other institutions directly related to stablecoins and thematic investments.
Mostaque, former head of Stability AI and now founder of Intelligent Internet, has long discussed open models and governance; Wood bets on AI, robotics, and Bitcoin through ARK thematic funds; and Diamandis has turned exponential technologies into sellable summit and podcast products. The conversation delved into the “agency will act” concept down to the levels of electricity, cabinets, and settlement, rather than just model rankings.
Healthcare was singled out due to regulations, liabilities, and data silos preventing agencies from directly replacing processes like customer service or programming. Humanoid robots were framed as a complex challenge involving hardware, batteries, fine motor skills, and on-site safety, which do not sync with the rapid iteration of software agents. The disparity in U.S.-China attitudes was used to explain deployment speeds, rather than new technological breakthroughs.
In market mechanisms, buyers are thematic funds and stablecoin issuers pricing the agency economy, while sellers are the hosts selling summit tickets, podcast ads, and infrastructure narratives. Funds flow from attention to energy projects, data center REITs, and on-chain infrastructure stocks. Beneficiaries include Circle and ARK's heavily invested targets that have integrated electricity and stablecoins into their product lines, while those under pressure are models still valuing AI based on software subscriptions without accounting for electricity costs and settlement delays. The event-driven nature comes from the packaged topics of this conversation, not from individual company earnings reports.
Source: Public Information
ABAB AI Insight
Diamandis's Moonshots turns the guest list itself into an asset: Mostaque provides open-source and governance rhetoric, Wood offers capital market multiples, and stablecoin sponsors provide the settlement layer. The agency economy is broken down into four investable shortages at this table—electricity, cabinets, dollar on-chain, and cross-border infrastructure—rather than the market cap of a single model company.
The capital path is narrative first, assets second. Energy and data centers are multi-year capital expenditures, stablecoins are existing settlement products, and robots are hardware that has yet to balance unit economics. The motivation is to rewrite "agents will replace jobs" into "first secure electricity and infrastructure." The strategy aligns with ARK's Big Ideas annual report and Circle-sponsored tech summits: first define the bottleneck, then make the targets in the portfolio the solutions to that bottleneck.
Similar cases include writing blockchain as all infrastructure in 2017, GPUs as the only bottleneck in 2023, and consistently framing healthcare AI as an exception that will only materialize in five years. The industry is in an expansion phase of the agency narrative: software agents can already demonstrate, while electricity and humanoids are still controlling who can mass-produce.
Structural judgments belong to the reconstruction of the industrial chain. Value is shifting from model weight to electricity, cooling, stablecoin settlement, and the robots themselves. The mechanism is: when agents need to truly move funds and objects, constraints shift from parameter quantities to grid capacity and settlement finality.
ABAB News · Cognitive Laws
The upper limit of the agency economy is first written on the electricity bill.
Stablecoin infrastructure is closer to settlement than model parameters.
Humanoid robots are slow because the world won't make way for software.