Chris Dixon on stablecoins hitting $2T monthly volume, crypto regulatory momentum, and the startup talent shortage

Summary

Chris Dixon, general partner at Andreessen Horowitz and one of the earliest institutional investors in Coinbase, argues that crypto is past its infrastructure bottleneck and into a talent shortage — the opposite problem from AI.

Stablecoins are the headline number. Dixon cites $2 trillion in monthly stablecoin volume, which he says now exceeds Visa. The underlying rails — Solana, Ethereum — can move any amount of money anywhere in the world for under one cent in under one second. That took years of investment to get right, but the infrastructure excuse is gone.

The more interesting argument is what comes next. Dixon frames stablecoins the way WhatsApp framed SMS: an over-the-top global network that sidesteps the patchwork of 195 national banking systems. Current use cases — cross-border treasury management (he cites SpaceX using stablecoins after Starlink sales in Brazil), remittances, and currency substitution in volatile economies — are the early layer. The programmability layer is where Stripe is focused. Dixon points to the Stripe-Bridge acquisition ($1 billion) as the clearest signal that mainstream fintech has committed. Patrick and John Collison's pitch, per Dixon, is less about low fees and more about programmable reputation systems and fraud prevention that traditional wiring infrastructure can't support.

Longer term, Dixon expects stablecoins to become invisible infrastructure — "digital dollars" embedded in fintech stacks, AI agent payment rails, and machine-to-machine micropayments. He anticipates MCP gaining a payment standard in a future version, enabling AI agents to autonomously negotiate and transact in fractions of a cent.

Regulatory momentum is real but fragile. Dixon and Marc Andreessen have been making monthly trips to Washington for several years. His framing is that startups can't afford DC presence, so a16z goes as their proxy. The priorities: a federal stablecoin bill (a procedural Senate vote last week drew 17 Democratic votes), a market structure bill advancing in the House, open-source AI protections, and a single federal framework rather than 50 state regimes. Dixon draws an explicit parallel to Section 230 of the 1996 Telecom Act — the argument being that legislative clarity, not administrative guidance, is what lets industries build reliably over decades.

He pushes back on the Republican-only framing of crypto politics. Campaign spending from crypto-aligned donors was roughly 50/50 across parties, and he describes the goal as pulling Democrats back toward the pro-innovation posture of the Clinton and Obama eras.

Real-world assets are the next frontier. Dixon treats stablecoins as the first RWA — dollars on-chain — and sees Robinhood's equity tokenization and BlackRock's on-chain Treasury bills as the natural next step. The logic he finds most compelling for institutional adoption isn't the fee reduction but the coordination problem blockchains solve. Banks won't build shared bond-trading infrastructure because Goldman won't trust a JP Morgan-run network, and neither trusts a startup that might raise fees later. A credibly neutral blockchain removes that political deadlock. He describes a portfolio company called Story Protocol that is putting intellectual property on-chain to handle licensing and capital formation in a world where AI can remix any creative asset.

The investment structure a16z uses for crypto has evolved. The firm runs a dedicated crypto fund, launched around 2018, with maximum flexibility to hold Bitcoin directly, make pure equity bets like Coinbase, or use an equity-plus-token-warrant structure it helped pioneer. That third structure gives founders the option to stay as a traditional equity company or launch a token later, with investors participating alongside them either way. Dixon argues the older SAFT structure misaligned incentives by pushing founders toward unnecessary token launches; the warrant model ties investor and founder outcomes together regardless of which path the company takes.

The Uniswap example illustrates what happens post-token. The C-corp became one front-end provider on a decentralized protocol doing trillions in volume; the company's website now accounts for under 5% of protocol volume but that is still a viable business.

The talent shortage is the actual constraint. Dixon's clearest pitch to listeners: crypto is undercompeted. The highest-usage crypto applications top out around 50 million users — roughly 1% of the internet's 5 billion people. For every good crypto idea, there are far fewer startups pursuing it than in AI, where Dixon estimates 50 companies chase every viable concept. His ask is direct: smart people considering a startup should look at crypto, where white space is abundant and the infrastructure is finally ready.