Matt Hougan, chief investment officer at Bitwise Asset Management, a US-based crypto asset manager and issuer of spot Bitcoin and Ethereum ETFs, said Wednesday that investors are significantly underestimating how much transaction activity blockchain networks will handle as autonomous AI software and tokenized financial assets converge on the same rails. In comments published by The Block on August 19, Hougan put a rough range on the gap: transaction volume could run 10 to 100 times higher than current models assume.

The argument rests on two structural shifts happening simultaneously. First, AI agents, autonomous software programs that can hold crypto wallets, pay for services, manage lending positions, and settle trades without human sign-off on each action, are already generating measurable on-chain volume. Second, traditional financial assets are being moved onto blockchains at an accelerating pace, creating a growing pool of assets that software programs can transact with around the clock.

Hougan frames the next phase of crypto growth as being driven by stablecoins, instant settlement, 24/7 trading, and institutional participation in decentralized finance, rather than speculative retail cycles. He also pushes back against a popular bearish view that blockchains will eventually become commodity infrastructure with collapsing fee revenue (a view sometimes called the "commodity blockspace thesis"). His counter is direct: demand from AI agents and tokenized markets will grow faster than supply. Blockchain throughput has already scaled roughly 100x over five years, rising from around 25 transactions per second to approximately 3,400 TPS, yet demand has kept pace.

The on-chain numbers, while still early, are not trivial. Coinbase's x402 protocol, a machine-to-machine payment standard that lets software programs pay each other without human intermediaries, launched in May 2025 and had processed more than 165 million transactions across roughly 69,000 active agents by April 2026. In May 2026 alone the protocol logged 3.1 million transactions worth about $1.2 million in value. Solana processed 10 billion total transactions in the first quarter of 2026 and currently processes around 65 percent of agentic AI payments running through the x402 standard. BNB Chain reported more than 150,000 registered autonomous entities by April. A first-half 2026 estimate put total economic output attributable to on-chain agents at roughly $470 million, according to one industry estimate.

On the tokenized asset side, the picture is larger but also more uneven. On-chain real-world assets (RWAs), a category that includes tokenized treasury bonds, private credit, equities, and real estate, reached $32.2 billion in total value by the end of June 2026, up from about $11.8 billion a year earlier. A broader Forbes accounting put the figure closer to $60 billion by July. Tokenized equities specifically crossed $963 million in January 2026, a roughly 2,878 percent increase year-over-year. The critical caveat is that most of that capital is sitting idle. Of the $60 billion tokenized, only about $3.97 billion is actively deployed within DeFi protocols. Hougan has argued that tokenization could expand the addressable market for crypto applications from approximately $2 trillion to $500 trillion. If his thesis plays out, AI agents could serve as the connective tissue between static tokenized assets and active financial markets, closing the gap between capital parked on-chain and capital put to work.

Coinbase CEO Brian Armstrong has made a similar case from a product standpoint. Armstrong stated on July 26, 2026 that AI agents "can't open a bank account, but they can own a crypto wallet," and predicted they will eventually execute more daily transactions than human users. Vibhu Norby, who oversees product strategy and AI adoption at the Solana Foundation and whose organization has a direct commercial stake in high Solana transaction volumes, put it more sharply: "99.99% of all onchain transactions in 2 years will be driven by agents, bots, and LLM-based wallets."

For users outside the United States, particularly in high-remittance markets, the stakes in this debate are concrete. Nigeria ranks second globally on Chainalysis's crypto adoption index and accounts for roughly 60 percent of stablecoin inflows across sub-Saharan Africa since 2019. Nigerian businesses paying overseas suppliers currently absorb fees of 6 to 10 percent through conventional banking; stablecoin transfers typically cost 2 to 3 percent and settle in minutes. If AI agents begin autonomously routing cross-border payments and managing treasury positions for small businesses, as Hougan's scenario implies, merchants in Lagos or Kano would benefit without needing deep technical expertise. India, which ranks first on the same global adoption index and processed an estimated $89 billion in stablecoin volume from Indian addresses in 2024, has a large developer base in Bangalore, Hyderabad, and Pune already building on tools such as ElizaOS, an open-source agent framework with more than 17,600 GitHub stars, and Coinbase's AgentKit. Pakistan, another high-volume remittance corridor where SWIFT transfer costs routinely exceed 3 percent, has also seen growing media and investor interest in tokenized asset research, signaling expanding regional engagement with the thesis. Regulatory friction remains real across all three markets: the Reserve Bank of India restricts private stablecoins in favor of its digital rupee pilot, and Nigeria's SEC is still working out compliance rules for autonomous transaction attribution.

Hougan's broader macro read is that 2026 fundamentals are building the case for a breakout in 2027. Whether the 10x to 100x transaction range proves accurate is unknowable at this point. What the data does confirm is that AI agents are already a measurable and growing source of on-chain activity, and that the tokenized asset market has scaled far ahead of actual usage, leaving a large gap that something will eventually have to fill. That question is moving beyond theoretical debate: the Africa Blockchain Festival 2026 in Nairobi this October and the Blockchain Africa Conference in Johannesburg both feature dedicated sessions on AI and blockchain convergence, reflecting how quickly this thesis is reaching applied-policy and developer audiences across emerging markets.