One thing to note is that Tether and Circle did not become incumbents because of branding or marketing, but because they each captured a distinct velocity layer as it emerged Tether won the remittance (money out of China) and offshore collateral markets (perps), helped along by Tron rails and the fact that much of global crypto demand still comes from places where simple dollar mobility matters more than anything else. USDC won the DeFi settlement layer largely by default, since Tether never invested in that ecosystem and Circle was willing to do the integrations in 2019-2021, along with strong offramp partnership with CB Both of these wins were tied to very specific economic flows. They were not timeless.. but were diachronic? IT was kinda mapped to the payment and leverage patterns of the era(s), and the stale capital that came along with too-fast-money doing nothing because the velocity was too quick Which is why the current scramble for the remaining 15% feels sorta directionless. People are fighting over the scraps of yesterday’s use cases rather than asking what the next velocity layer will even look like If you take a forward view, the next frontier is fairly obvious: agentic payments, even though its riffled with scams.. because agentic commerce isnt a thing yet. Right now its subscriptions. But this is a sector that probably far bigger than any crypto stale capital. Agents will transact continuously, autonomously, and with predictable spending patterns based on compute consumption, the first greenfield stablecoin market in years... but which currency? Why would it use USDC or USDT? It is unlikely to be off-ramps or "brand." Those are human. x402 is machine to machine. Agents will optimize around the currency that aligns with their operating costs and constraints. And today, those constraints are dominated by subscriptions tied to compute usage.. but whatever subscriptions will settle in, will be what they probably evolve from for agentic/per-action compute payments. Compute pricing flows directly from the financing of the underlying hardware. AI capex is already the highest velocity of capital seen in the US today. The thing is interest rates, namely the cost of compute-tied debt, is setting the floor for everything above it. Nvidia looks at the sustainable debt rates to price their machines, which are then sold to data centers, which are then rented out to ai companies, who are then rented out to end users, the debt controls the equity as always. This creates an unusual dynamic. The stablecoin that can return value back into this cycle (by lowering operator financing costs when payments flow through it) becomes economically sensible for agents to adopt. Not because it is marketed well, but because it reduces the cost structure of the systems they rely on. Much like using dollars.. because thats what we pay our mortgage and taxes with. Petrodollar 2.0