All takes are our summaries. Tap View on Xfor the analyst's original words.
Thesis: compute becomes a commodity, like oil. Supply-constrained today, but heading toward standardization. Like oil, it needs market infrastructure — futures, storage/logistics, price discovery, hedging instruments. First movers are the cloud operators; the real prize is the exchange layer that gets built atop them. The venture opportunity is backing that layer, not the underlying chips or data centers.
Robbie argues the agentic commerce thesis is inverted: while the agentic economy will be massive, most agents won't transact autonomously. Commercial agents (95%+ of agentic deployment) embedded in SaaS won't spend money—they'll research, review, or generate output. Consumer agents will remain orchestrators requesting authorization, not independent economic actors. Only bottom-up agents outside organizational control genuinely need granular, autonomous payments, where blockchains' permissionlessness beats card networks' compliance friction. The real bottleneck isn't payment rails but regulatory frameworks and legal structures enabling autonomous decision-making—a protocol upgrade can't solve that. Most agentic economy activity gets billed monthly, not settled per transaction.