Author: Christy Choi
Compiled by: Deep Tide TechFlow
Deep Tide Introduction: Christy Choi was a core executive in the early days of Binance Labs, with ten years of experience in the crypto industry, and now manages a fund spanning Asia, the Middle East, and the United States.
In this article, she presents a judgment: the crypto industry is undergoing a fundamental transformation, driven by two simultaneous structural forces—institutional capital entering through stablecoins, and AI driving the cost of building everything down to nearly zero.
She believes that the era of token speculation is coming to an end, and the winners of the next era will be licensed infrastructure companies with real revenue that can support AI agents. Her views are sharp and worth reading against the current state of the market.
Full text as follows:
Something has changed in the past twelve months, but most people in the market haven’t realized it yet.
I’ve been in the crypto industry for ten years—serving as a core executive in the early days of Binance Labs, making early-stage investments and building projects across multiple cycles, and now managing a fund with a presence in Asia, the Middle East, and the United States. I’ve experienced every version of this industry: the ICO frenzy, DeFi summer, NFT bubble, and serial collapses. Each cycle felt different at the time, but they were all powered by the same engine—speculative capital chasing narrative tokens.
That engine is dying. Not because crypto has failed, but because it has succeeded. What replaces it will reshape this industry more fundamentally than anything since Bitcoin.
Two structural forces are converging at the same time: institutional capital entering through stablecoins, and AI driving the cost of building in crypto down to nearly zero. Together, they are changing not just which tokens will win, but what crypto itself is.
Institutional Shift: Stablecoins Devour Everything
Most crypto-native players haven’t internalized this: the largest wave of capital ever to enter this industry won’t buy tokens—it will use stablecoins.
Stablecoins now settle trillions of dollars annually. They are the first crypto product that institutions, corporations, and governments actually want—not for speculation, but as infrastructure. When a multinational company runs treasury operations through stablecoin rails, when a remittance corridor switches from SWIFT to USDC, when a neobank offers stablecoin-denominated savings accounts to the unbanked in Southeast Asia—these are real economic activities moving on-chain. Not TVL farming, not governance token speculation, but revenue.
This changes the entire value chain. The winners in this new landscape aren’t protocols with clever tokenomics, but licensed businesses with regulatory moats. Stablecoin issuers, compliance middleware providers, licensed neobanks, settlement infrastructure—these companies are capturing the largest share of the institutional wave. They are boring by Crypto Twitter standards, but they will generate the most durable returns of the next decade.
The moat for these businesses isn’t technology—it’s regulation.
This is where crypto-native players have consistently underestimated. In the old crypto world, moats came from liquidity, network effects, and community. In the new crypto world, the deepest moat is a license. Every jurisdiction that finalizes stablecoin rules, tokenization frameworks, or digital asset banking regulations creates a window—typically 12 to 18 months—where the first licensed operators build advantages that later entrants can’t replicate no matter how much money they spend. Customer relationships, banking partnerships, compliance infrastructure, regulatory trust—these things can’t be forked, and your agent can’t code them. By the time competitors get licensed in the same jurisdiction, the first movers have locked up distribution channels.
This happens jurisdiction by jurisdiction, not globally. Europe’s MiCA, emerging stablecoin frameworks in Singapore and the UAE, South Korea’s Digital Asset Basic Act, and the regulatory architecture now taking shape in Washington—each creates different license moats in different markets. The companies that win are those that treat regulation as a strategic asset, not an obstacle. They are hiring former regulators, not just engineers. They are shaping frameworks, not just complying. Policy proximity—the ability to influence rulemaking while rules are being written—is the most valuable and least understood competitive advantage in crypto right now.
The native token playbook—launch token, attract TVL, pump the price with narratives, cash out through unlocks—is ending because the capital entering the system doesn’t play that game. Institutional allocators want yield, compliance, and predictable revenue. They don’t want governance rights to a protocol that might be forked next quarter.
As real cash flow moves on-chain, tokens inevitably become more like equity. When a protocol has real revenue and distributes fees to token holders, the token is no longer a speculative instrument—it becomes a machine-readable ownership claim on a real business. This is convergence. Not tokens replacing equity, nor equity replacing tokens, but both collapsing into the same thing: programmable, composable, instant-settling claims on real economic activity. The wrapper no longer matters; what matters is the underlying business generating cash, and the rights to it being software-readable.
The same license moats apply to tokenization platforms. When stocks, bonds, and structured products move on-chain, the platforms tokenizing them won’t be permissionless protocols—they will be licensed securities intermediaries operating under specific regulatory frameworks in specific jurisdictions. The infrastructure is crypto-native, the business model is TradFi-grade, and the moat is still licenses, not code.
Crypto spent a decade building rails to move value. But the rails to move identity, professional credentials, and authorization haven’t been built. This is the gap—and where the next wave of infrastructure will be built. Teams that figure out attested identity, machine-verifiable compliance, and portable professional credentials at the on-chain primitive level are building the layer that connects institutional finance and autonomous AI. Neither side works well without it.
AI Shift: Building Becomes Cheap, Verification Becomes Precious
The second force is AI, and its impact on crypto is deeper than the “AI × crypto” narrative suggests.
First, the obvious: AI drives the cost of building any software to extremely low levels. Spinning up an L2, deploying a set of smart contracts, launching a DeFi primitive—all of this can now be done in days, with engineering teams a fraction of the size they used to be. The implications for existing infrastructure are brutal: when supply approaches infinity, premiums evaporate. The 100+ blockchains already live will be compressed to utility-level margins. Infrastructure that once commanded $1 billion to $5 billion FDV will be repriced to what it actually earns. The venture capital playbook of “invest in infrastructure, capture narrative premium, exit to retail” is structurally broken.
But AI does something else entirely, and this is where crypto goes from optional to necessary.
When AI agents can generate infinite transactions, content, identities, and interactions, the cost of faking anything approaches zero. Spam and valid signal become indistinguishable; bot activity and human activity become indistinguishable. In a world flooded with infinite machine-generated noise, the only way to establish trust is through cryptographic proof.
Zero-knowledge technology goes from a niche scaling solution to essential infrastructure. Privacy-preserving credentials go from academic research to the attestation layer of every AI-participating system. If you can’t prove who you are, that your transactions are authorized, and that your agent is credentialed without exposing the underlying data, you can’t participate.
This is the most underrated proposition in crypto right now: ZK and privacy tech aren’t privacy plays—they are the trust layer for the AI economy.
Where the Two Forces Converge: Tokens as the Machine Operating Layer
Where institutions and AI converge is where I believe the deepest opportunities lie.
AI agents are already trading autonomously. Coinbase just launched a wallet designed specifically for AI agents. The x402 protocol enables machine-to-machine payments. Autonomous systems are starting to hold assets, execute trades, pay for compute, and interact with financial services without human intervention.
These agents need three things to operate.
Identity. Not a username—a cryptographic credential that a counterparty can verify in milliseconds without seeing the underlying data. Who does this agent represent? In which jurisdiction does it operate? What is it authorized to do? If you can’t answer these questions with programmable proofs, you fall back to centralized databases and manual reviews. That doesn’t scale when millions of agents are trading simultaneously.
Programmable assets. Stablecoins proved that money can be programmed and settled instantly. The same logic extends to treasuries, stocks, credit, structured products. An agent doesn’t care if it holds USDC or tokenized treasuries. It cares that the terms are readable, the rules are programmable, and settlement is deterministic.
Credentials. Today, compliance lives in human judgment and legal documents. It needs to encode regulation into machine-verifiable proofs—KYC status, licenses, jurisdictional permissions, risk limits. The only way to deliver these at scale is cryptographic proofs, not centralized APIs.
This is where the two shifts I’ve been describing truly collide. Institutions push financial assets on-chain because programmable infrastructure reduces settlement risk and operational friction. AI pushes economic activity toward autonomous execution. When these two forces meet, the financial objects themselves must become software.
Machines won’t buy tokens to speculate. They consume tokens to operate. This creates a demand curve completely different from anything in crypto history. Retail speculation is cyclical, narrative-driven capital that rotates. Machine consumption scales with the volume of autonomous economic activity. As AI systems automate more decisions, trades, procurement, and coordination, demand for machine-readable financial objects scales in lockstep.
What This Means for Our Investments
I’m writing this as abstract analysis. This is the direction my new fund is actively investing in.
The old crypto game was: find the narrative, front-run the token, exit before unlocks. The new game is: find the licensed infrastructure layers capturing stablecoin flow, build the machine-readable primitives agents need to transact, invest in jurisdictions where regulatory frameworks are crystallizing first.
The companies that win the next era look completely different from the projects that defined the last one. They have licenses, not just liquidity. They have revenue, not just TVL. They have regulatory moats, not just network effects. By the standards of those who made their fortunes trading memecoins, they are boring. By the standards of those allocating institutional capital, they are a generational opportunity.
The token speculation era gave crypto its beginning. Institutions and AI will give it its future. The transition between the two eras is happening right now, and it’s moving faster than most participants realize.
The speed of this shift is the story no one has fully told yet. Consider this a first draft.
