Source: Bankless
Compiled by: Felix, PANews
Aleks Larsen and Spencer Bogart, General Partners at Blockchain Capital, recently appeared on the "Bankless" podcast to discuss the inevitable shift in the crypto market from infrastructure to the application layer. Blockchain Capital noted that the widespread adoption of stablecoins has already accumulated immense liquidity for on-chain finance, driving revenue growth for lending and trading protocols.
Furthermore, by tokenizing stocks and venture capital funds, the capital efficiency of the financial system will be exponentially improved. While there is a tension between traditional finance and crypto's ethos regarding compliance, the reconstruction of the global financial system through tokenization technology is unstoppable. PANews has compiled the highlights of the conversation.
Host: Spencer, I recall our first exchange in the industry was back in 2018 or 2019, discussing MKR's value capture model.
Spencer: Yes, we were even considering buying MKR at the time. Now, in 2026, the most modern projects like Hyperliquid, Lighter, and Venice are still adopting the "buyback and burn" model pioneered by MKR. Despite countless past debates about the low capital efficiency of this model, it has arguably been "undefeated" in practice.
Aleks: Indeed. I used to be very critical of this model, believing that in the endgame, when only the last token remains, there must be substantial cash flow that can be directly distributed to holders; otherwise, it's hard to build a valuation model. But now I think that was overthinking it. The "buyback and burn" model works very well today.
Spencer: Exactly. The main reason is that, unless the Clarity Act is passed, the legal rights of token holders remain very ambiguous. Theoretically, as an investor, if you're a startup, I would want you to reinvest cash flow into new growth opportunities. But in reality, most crypto protocols haven't demonstrated the ability to expand across domains and succeed. Therefore, many token holders prefer the team to "plant a flag in the sand," clearly signaling to the market, "We will forever buy back and burn," which at least eliminates uncertainty.
Additionally, due to the uneven quality of early crypto tokens. Serious, high-quality projects must use "real money" to buy back and burn from day one to prove they are different. While this may not be the dominant model in 5 years, at this stage, it's the most effective, credible way to align interests with token holders.
Host: There's a current narrative that "crypto VC is dead." All the big funds are expanding their investment scope to frontier tech like AI and robotics, but you at Blockchain Capital are doubling down during the industry downturn. Strangely, I see two extremes simultaneously: on one hand, traditional financial institutions are eager to get into blockchain; on the other, crypto OGs are very pessimistic. How should we understand this rift?
Aleks: We are accustomed to zooming out and not overly focusing on price fluctuations during bull/bear cycles. This "token bear market" is actually very special because it's accompanied by the most positive catalysts ever. We've welcomed the Genius Act and the gradually clarifying Clarity Act. Rules are being established, and traditional institutions are entering en masse.
More importantly, some applications have broken through the industry's information silo and entered the mainstream: such as prediction markets (projects like Polymarket, where many users don't even care if it's crypto-based) and stablecoins (providing extremely cheap cross-border dollar payments and remittance channels). These sectors have still achieved strong one-sided growth during the bear market. It's just that over the past year or so, AI has absorbed all the market's attention, especially the explosion of coding agents and open-source Claude 7-8 months ago, causing many to lose focus when token prices were low.
Host: You often mention the "S-curve." Can you explain in detail where the crypto industry is on this curve now?
Aleks: The crypto industry's development trajectory is highly similar to the internet. The internet commercialized starting in 1989. For the first 10 years, it was exploratory. By 2000, there were hundreds of millions of users, but it was still extremely difficult to use and bandwidth-limited. Then came the broadband transition from 2000-2005. I believe the crypto industry has just undergone its own "broadband transition." Block space has become extremely cheap and abundant. In 2020, Solana was the first monolithic chain to demonstrate high performance and a path to scaling. By 2024, L2s truly became widespread, and even Ethereum is gradually achieving scaling. This has become the new normal for the industry.
Looking back at the internet, after the broadband transition was complete, it didn't explode immediately. It wasn't until the mobile explosion of 2006-2010 that the S-curve bent upward. If the launch of Ethereum in 2015 represents the starting point of the "clock," we are only 10 to 11 years in development. Among the 700 million crypto holders, perhaps only 10% are active on-chain users. It's only in the last 2-3 years that truly user-friendly, consumer-grade technology stacks (like embedded wallets, social recovery, spending limits, and passwordless login) that don't require users to be cryptographers have matured and become widespread.
Therefore, we are currently in the 2003-2004 phase of the internet, the "flat bottom of the S-curve" after broadband penetration but before the mobile explosion. Once edge applications like stablecoins and prediction markets fully penetrate the mainstream, the S-curve will experience an upward inflection point.
Host: Perhaps our generation was too young and impatient in 2021, thinking we could change the world tomorrow. Actually, technology and infrastructure need time to mature. But this still doesn't fully explain why the OGs are so frustrated.
Spencer: It's psychological "growing pains." When a startup reaches the IPO stage, early core employees often miss the days of being "rebellious pirates" and struggle to accept the company transforming into a compliant, massive entity to succeed. It's like when a friend discovers a very niche, unique band, but when that band blows up and becomes mainstream, he feels regret, declaring, "I only like their early albums."
Aleks: Yes, industry conferences are now full of people in suits, talking about permissioned channels, compliance, and access, not cypherpunk. But finance is inherently a highly regulated sector; you can't scale without playing by the rules.
However, the decentralization and neutrality of Ethereum and Bitcoin still hold immense underlying appeal for institutions because they offer better trust assumptions. The cypherpunk dream hasn't died; it's just operating in a quieter, more scaled form as the underlying network for the financial system. We are genuinely upgrading the plumbing of the global financial system. While it may sound less "sexy" than the old days, the efficiency gains will tangibly benefit everyone.
Host: Indeed. And you mentioned a detail earlier: For the first time in history, traditional institutions are actively diving deep and deploying into crypto assets during a price decline, without market frenzy narratives. Also, in 2025 and 2026, the industry seems to have completely broken the cycle of "investing in infrastructure for infrastructure's sake." What does this represent in terms of industry evolution?
Spencer: In 2019, interacting with Uniswap could cost a few dollars or even over ten dollars in friction costs. Back then, severely insufficient block space was the biggest bottleneck in the industry. This led to an over-investment of capital in infrastructure driven by market frenzy, resulting in today's situation of a severe oversupply of block space and many blocks being empty. But abundant, cheap block space is an absolute prerequisite for application developers to thrive.
The data is very clear. In 2021, over 70% of fees paid by users went to the infrastructure layer. In 2025, the total fees at the application layer historically surpassed the infrastructure layer for the first time. This means that as transaction costs plummet, value is finally shifting up the protocol stack to the application layer. A healthy ecosystem shouldn't allow the underlying communication infrastructure to extract the vast majority of monopoly rents. This is precisely the traditional banking rent model that crypto technology seeks to break.
Host: So, this is the so-called "Fat Application Theory" replacing the earlier "Fat Protocol Theory"?
Aleks: Exactly. The underlying protocol layer shouldn't retain massive profits because the nature of blockchain is to reduce intermediary rent-taking and improve efficiency. But the more advanced logic is "thin protocol, large market": Even if your take rate is extremely low, once you expand the underlying market size of global finance by an order of magnitude, the total absolute value captured will still be enormous.
Host: That's interesting. If we extrapolate this "Fat Protocol to Fat Application" shift to the AI field, does AI investment and evolution follow similar patterns?
Aleks: The similarities are striking. In the crypto industry, teams could raise valuations of tens of billions of dollars with just a whitepaper, much like how AI labs today easily secure sky-high valuations with research visions and star teams. In crypto, we look at testnet TPS and benchmarks; in AI, it's various model benchmarks. In crypto, exchange listings provide liquidity; in AI, it's gaining distribution channels through hyperscale cloud providers.
But there's one huge difference. In crypto, token prices are a completely public and transparent sentiment thermometer; once a narrative breaks, tokens can plummet 90% in a month. The bubble and downward pressure in the AI field are currently hidden in the private capital markets. It may not crash directly like crypto but manifest as down rounds, talent drain, etc.
Host: Will the application layer in AI explode similarly to crypto?
Spencer: Absolutely. As Palantir Technologies CEO Alexander emphasizes, having models and intelligence alone doesn't directly produce the outcomes enterprises want. Someone must go to the front lines to convert that intelligence into actual workflows and outputs.
Interestingly, AI VCs have recently been gripped by panic over "software having no moat." As crypto VCs, we find this amusing because for the past 10 years, the crypto industry has dealt daily with a brutal environment of "everything is open source, anyone can fork the code at any time, with no software moat whatsoever."
Aleks: General model weights will gradually commoditize, but the "harness" that utilizes them to solve real-world problems won't. In complex, hardcore domains where "no error is tolerated" (like semiconductor manufacturing, complex tax audits, etc.), applications that leverage frontier models plus fine-tuning techniques, supplemented by proprietary enterprise datasets and closed-loop feedback, will build incredibly deep moats that generic models cannot breach.
Host: Back to RWA tokenization. As the first generation of the most successful RWAs, what lessons does stablecoin development offer us?
Spencer: Few people know that Blockchain Capital is the only venture firm that invested in all three major stablecoin issuers (Tether, Circle, Paxos) a decade ago. The total stablecoin market cap today is around $300 billion. I am almost 90%+ confident that by 2030, this number will soar to several trillion dollars (even $2 trillion). Previously, stablecoins were driven by a retail flywheel. But now, every new flywheel turning is accompanied by institutional push, bringing traditional stocks, money market funds, and treasury bonds "on-chain," because the capital efficiency of a globally 24/7, programmable underlying network is just too high.
The core of a stablecoin is far more than just a "payment product." Its stickiness is extremely high. Once dollars are on-chain, the vast majority of that capital settles and is injected as working capital into lending, exchanges, and other on-chain ecosystems, catalyzing massive economic activity. We've conducted precise quantitative measurements: Every $1 billion in net new stablecoin issuance creates approximately $122 billion in economic activity on-chain over a year. That $1 billion will directly generate about $19 million in recurring protocol revenue for downstream on-chain protocols within a year.
Host: So, besides stablecoins, how will the highly anticipated "stock tokenization" evolve?
Spencer: Stock tokenization has two waves. The first wave is access. Global investors (especially non-US users) have a strong demand for frictionless, one-click trading of US stocks. The second wave is composability. Once my Apple stock token is on-chain, countless lending services and securities lending protocols can compete openly in the market to offer me the best collateral rates and yields. This is the ultimate expression of capital efficiency.
Currently, there are two main competing approaches in the market. One is the X-Stocks model represented by Backed (acquired by Kraken). It issues debt instruments via Cayman SPVs to peg to stocks. The advantage is that it's completely permissionless, no KYC required, and can freely circulate in DeFi. The fatal drawback is that you own the debt owed to you by the SPV, not the actual Apple stock share. For large institutions with tens of billions in capital, this credit and legal risk is unacceptable. The other is a compliant channel that directly owns stock ownership. This requires compromise on permissionlessness.
Host: So, does this mean the "suit-wearing bigwigs" of traditional finance and the "pirates" of crypto must make one side compromise?
Spencer: Not necessarily. We don't have to force them to merge. Those trillions in traditional stocks can run in "sidecar mode" on the main public chain. They may have regulatory fences but exist alongside pure, permissionless DeFi liquidity pools. This could actually greatly accelerate liquidity in pure cypherpunk systems because the massive capital parked in stock tokens can be converted into ETH with one click and operate in purely decentralized, permissionless environments.
Related read: Conversation with the 'Liquidity King': Global Liquidity Has Peaked and Is Declining, This Cycle Will Bottom in the Second Half of Next Year
