Blockchain Capital general partners Aleks Larsen and Spencer Bogart said on a recent Bankless podcast that crypto is moving from an infrastructure-led market to one centered on applications. In their view, stablecoins, prediction markets and tokenized financial assets are the clearest signs of that shift.
The conversation, summarized by PANews and sourced to Bankless, also laid out a wider thesis: cheaper block space is changing where value accrues, application fees are overtaking infrastructure fees, and traditional financial institutions are entering the sector even without the kind of euphoric price backdrop seen in earlier cycles.
Buyback-and-burn is still working
The host opened by recalling an earlier conversation with Bogart in 2018 or 2019 about Maker’s MKR value-capture model.
Bogart said they had even considered buying MKR at the time. He argued that by 2026, newer projects such as Hyperliquid, Lighter and Venice were still using the buyback-and-burn structure first popularized by MKR. That model has been criticized for years as capital-inefficient, he said, but in practice it has remained remarkably resilient.
Larsen said he used to be far more skeptical. His older view was that any token model ultimately needed real cash flow distributed directly to holders if investors were supposed to value the last unit of supply. He now sees that line of thinking as too theoretical for the current market. For now, he said, buyback-and-burn works.
Bogart tied that back to regulation. Unless the Clarity Act passes, token holders’ legal rights remain unclear, he said. In a traditional startup, an investor may prefer management to reinvest cash flow into new growth opportunities. In crypto, though, many protocols have not shown they can expand outside their original lane and do it well. In that setting, token holders often prefer a simple, durable commitment: keep buying back tokens and destroying them.
He also said the uneven quality of early crypto assets shaped market expectations. Serious teams have had to show from day one that they are different, and using actual revenue to buy back and burn tokens has become one of the most credible ways to do that. Bogart added that the model may not dominate five years from now, but he sees it as one of the most effective forms of alignment at this stage.
A token bear market with positive catalysts
The host then raised a common market claim that “crypto VC is dead,” noting that many large funds have broadened their remit to include AI, robotics and other frontier sectors. Blockchain Capital, by contrast, chose to lean in while the sector was weak. At the same time, established financial institutions appear eager to get involved with blockchain, while many crypto OGs sound deeply pessimistic.
Larsen said his firm tries to zoom out instead of focusing too heavily on price action across bull and bear cycles. He described the current token downturn as unusual because it is happening alongside what he called the strongest set of positive catalysts the industry has seen. He pointed to the Genius Act, a gradually clearer path for the Clarity Act, and a rules framework that is starting to take shape as institutions come in at scale.
He also argued that some products have already broken out of crypto’s internal bubble. Prediction markets were one example, with Polymarket cited specifically. Many users do not care whether the backend is “crypto” in ideological terms, he said. Stablecoins were the other major example because they offer very cheap dollar-based cross-border payments and remittances. According to Larsen, both sectors have posted strong one-way growth even during the bear market.
He said AI has absorbed much of the market’s attention over the past year or more, especially after coding agents and open-source Claude accelerated seven or eight months ago. That left many market participants distracted while token prices stayed weak.
Crypto looks like the internet in 2003 or 2004
Asked to explain the S-curve framework that he often uses, Larsen compared crypto’s development path with the internet’s commercial rollout.
He said the internet began commercializing in 1989 and spent its first 10 years in an experimental phase. By 2000, it had reached hundreds of millions of users, but it was still difficult to use and constrained by bandwidth. The period from 2000 to 2005 then brought the broadband transition.
Larsen said crypto has just been through its own version of that transition. Block space has become cheap and plentiful. In his account, Solana was the first chain to show a viable high-performance monolithic scaling path in 2020. By 2024, Layer 2 networks had started scaling rapidly in real terms, and Ethereum itself was also moving toward broader scalability. He described that as the new normal for the industry.
Still, the internet did not jump immediately after broadband improved the base layer. Larsen said the steeper part of the S-curve came later, from 2006 to 2010, when mobile usage exploded. If Ethereum’s launch in 2015 is treated as the starting clock for modern crypto, he said, the sector is only 10 to 11 years into that process.
He added that out of roughly 700 million crypto holders, perhaps only 10% are active on-chain users. The reason, in his view, is that genuinely usable consumer-facing tools only became mature over the last two or three years. He cited embedded wallets, social recovery, spending controls and passwordless login as examples.
That leaves crypto, in Larsen’s framework, at the equivalent of the internet in 2003 or 2004: after broadband, before mobile, and still on the flat lower section of the S-curve. Once edge applications like stablecoins and prediction markets move closer to the center, he said, the curve could start bending upward.
Why many OGs feel disillusioned
The host suggested that many people in crypto may simply have been too young and impatient in 2021, expecting the world to change overnight. Even so, that alone does not explain the disappointment visible among many long-time participants.
Bogart called it a kind of growing pain. When a startup gets close to an IPO, he said, early employees often miss the period when the company felt like a rebellious pirate ship and struggle with the idea that success requires becoming a large, compliant organization. He compared it to someone who discovers a niche band early, only to lose interest once that band becomes mainstream and starts filling arenas.
Larsen said the same shift is visible at industry conferences, where discussions now revolve around permissioning, compliance and access rather than cypherpunk ideals. Finance is a heavily regulated domain by nature, he said, and any attempt to operate at large scale has to engage with that reality.
At the same time, he argued that Ethereum and Bitcoin still hold deep appeal for institutions because of their decentralization and neutrality. Those systems offer stronger trust assumptions, in his view. The cypherpunk vision has not disappeared; it has moved into the background and is operating at greater scale as the base network for a financial system being rebuilt. That may be less romantic than earlier crypto narratives, he said, but the efficiency gains are tangible.
Applications overtook infrastructure in 2025
The host then returned to one of Blockchain Capital’s recurring points: for the first time, institutions are moving into crypto during a period of falling prices and without a dominant speculative mania. At the same time, 2025 and 2026 appear to mark the end of a cycle in which the market kept funding infrastructure for infrastructure’s sake.
Bogart said that in 2019, a simple interaction on Uniswap could cost users several dollars or even more than $10 in friction. Block space scarcity was the sector’s main bottleneck. That bottleneck pulled too much capital into infrastructure during the market’s speculative phase, and the result today is a large oversupply of block space, with many blocks sitting partly empty.
He was clear, though, that abundant and cheap block space is exactly what application builders need.
His data point was straightforward: in 2021, more than 70% of user-paid fees flowed to infrastructure. By 2025, total application-layer fees had exceeded infrastructure-layer fees for the first time. Bogart said that shift shows value finally moving up the stack as transaction costs collapse.
He added that a healthy ecosystem should not let the communications base layer extract most of the monopoly rent. Breaking the rent-seeking model common in traditional banking is one of the reasons crypto exists, he said.
From fat protocols to fat applications
That led to a direct question from the host: is the old “fat protocol” thesis giving way to a “fat application” thesis?
Larsen said yes. Protocols should not be retaining oversized profits if the purpose of blockchains is to reduce intermediation and improve efficiency. His broader formulation was “thin protocols, big markets.” Even if the take rate at the base layer is low, he said, the absolute value captured can still be very large once the underlying market expands by an order of magnitude.
Crypto and AI rhyme, but they are not identical
The conversation then shifted to AI. Asked whether the same transition from infrastructure to applications might also apply there, Larsen said the parallels are hard to miss.
In crypto, teams once raised multi-billion-dollar valuations with little more than a white paper. In AI, he said, a number of startup labs are doing something similar with research vision and elite talent. Crypto investors tracked testnet TPS and benchmarks; AI investors track model benchmarks. Listing on exchanges gave crypto projects distribution and liquidity; AI companies pursue distribution through hyperscale cloud providers.
But he stressed one major difference. Token prices in crypto are public and liquid, making them a visible sentiment gauge. When a narrative breaks, a token can lose 90% in a month. In AI, by contrast, the bubble and the downside are still largely held inside private capital markets. The unwind may show up not as a violent public repricing but through down rounds, talent departures and slower follow-on funding.
When asked whether AI applications will also break out, Bogart said yes without hesitation. He referred to remarks by Palantir Technologies CEO Alexander that models and intelligence alone do not produce enterprise outcomes. Someone still has to translate that intelligence into workflows, execution and real output.
He also said many AI venture investors have recently become anxious that “software has no moat,” a concern that looks almost ironic from a crypto perspective. Crypto investors, he said, have spent the past 10 years operating in an environment where everything is open source, anyone can fork the code at any time, and conventional software moats barely exist.
Larsen added that generic model weights are likely to become commoditized, but the harness built around them will not. In high-stakes domains where errors are unacceptable, such as semiconductor manufacturing or complex tax auditing, he said, applications that combine frontier models, fine-tuning, proprietary enterprise datasets and closed-loop feedback could develop moats that generic models cannot easily break.
Stablecoins as the first successful RWA category
The discussion then returned to tokenized real-world assets. The host asked what the rise of stablecoins, which he described as the first successful generation of RWAs, has taught the market.
Bogart said few people realize that Blockchain Capital was the only venture firm to invest a decade ago in all three major stablecoin issuers: Tether, Circle and Paxos.
He said stablecoins now have an aggregate market value of roughly $300 billion, and that he has more than 90% confidence that the figure will rise into the trillions by 2030, “even $2 trillion.” Earlier stablecoin growth was driven mainly by retail activity, he said. Now, each new growth flywheel is increasingly institution-led and is pulling traditional equities, money market funds and Treasuries on-chain because a 24/7 programmable network is more capital-efficient.
Bogart argued that the core of the stablecoin story is not just payments. Once dollars move on-chain, most of that capital tends to stay there. It becomes working capital for lending markets, exchanges and other on-chain venues, which then generates broader economic activity.
He said Blockchain Capital had modeled the effect in quantitative terms: every net $1 billion in new stablecoin issuance creates about $122 billion in on-chain economic activity over one year, while directly generating about $19 million in recurring protocol revenue for downstream on-chain applications over the same period.
Two waves for tokenized equities
On tokenized stocks, Bogart said the market is likely to develop in two stages.
The first is access. Global investors, especially those outside the US, have strong demand for a low-friction way to buy and trade US equities with one click. The second is composability. Once an Apple share exists as a token on-chain, a wide range of lenders and securities-lending protocols can compete in open markets to offer the best collateral terms and the best yield. That, he said, is where capital efficiency reaches its fullest expression.
He described two competing structures in the market today.
- One is the X-Stocks model associated with Backed, which he noted has been acquired by Kraken. In that model, a Cayman SPV issues debt instruments linked to the underlying stock. Its strengths are that it can operate without permission, without KYC, and can move freely through DeFi. Its weakness is fundamental: the holder owns a claim against an SPV, not an actual equity interest in Apple. For institutions allocating tens of billions of dollars, Bogart said, that legal and credit risk is unacceptable.
- The other route is a compliant channel in which investors directly hold ownership in the stock itself. That route, he said, requires giving up some of the permissionless qualities that crypto users prize.
Traditional finance and crypto do not have to merge into one system
The host closed by asking whether that means either Wall Street-style institutions or crypto-native “pirates” ultimately have to give ground.
Bogart said no. He argued that the two systems do not need to be forcibly merged. Traditional equity markets, worth tens of trillions of dollars, can run in what he called sidecars on public mainnets. They can remain fenced by regulation while existing alongside fully permissionless DeFi pools.
In his view, that setup could actually speed up liquidity growth for native cypherpunk systems. Capital parked in tokenized stocks could, at least in principle, be converted into ETH with one click and then move into fully decentralized, permissionless environments.
The broader conclusion from both investors was consistent throughout the interview: once block space is cheap, regulatory rules are clearer and institutions are starting to participate, value capture in crypto shifts upward. Stablecoins, prediction markets and tokenized real-world assets are the areas they see as most capable of driving the next phase.