Former Binance Labs executive: Under the impact of AI+institutions, the cryptocurrency industry will undergo significant changes in two years Author: Christy Choi Compiled: Deep Tide TechFlow Deep Tide Introduction: Christy Choi served as a core executive at Binance Labs in the early days of its establishment, and has been deeply involved in the cryptocurrency industry for ten years. Now, she manages a fund that spans across Asia, the Middle East, and the United States. She made a judgment in this article: the cryptocurrency industry is undergoing a fundamental transformation, driven by two structural forces simultaneously - institutional capital entering through stablecoins, and AI pushing the cost of building everything close to zero. She believes that the era of token hype is coming to an end, and the next winners will be infrastructure companies with licenses, real income, and the ability to serve AI agents. Clear viewpoints, suitable for reading against the current market situation. The full text is as follows: Something has changed in the past twelve months, but most people in the market have not yet reacted. I have been in the cryptocurrency industry for ten years - serving as a core executive in the early days of Binance Labs, conducting early-stage investments and project construction across multiple cycles, and now managing a fund with a presence in Asia, the Middle East, and the United States. I have experienced every version of this industry: ICO mania, DeFi summer, NFT foam, and chain thunderstorms. Each cycle feels different at the time, but behind it runs the same engine - speculative funds chasing narrative tokens. This engine is dying. Not because the encryption failed, but because it succeeded. The thing that replaces it will fundamentally reshape this industry more than anything since Bitcoin. Two structural forces are converging simultaneously: institutional capital entering the market through stablecoins, and AI pushing the cost of building everything encrypted to near zero. Together, they change not only which tokens win, but also what encryption itself is. Institutional shift: Stablecoins devour everything. Most native crypto players have not internalized this: The biggest wave of funds ever entering this industry will not buy tokens, it will use stablecoins. Stablecoins now settle trillions of dollars annually. It is the first encryption product that institutions, businesses, and governments truly want - not as speculation, but as infrastructure. When a multinational company operates fund management through stablecoin tracks, when a remittance channel switches from SWIFT to USDC, when a new bank offers stablecoin denominated savings accounts to Southeast Asian populations lacking banking services - these are real economic activities moving onto the chain. It's not TVL mining, it's not governance token speculation, it's revenue. This has changed the entire value chain. The winner in the new landscape is not the protocol with clever token economics, but the licensed enterprise with regulatory moats. Stablecoin issuers, compliance middleware providers, licensed new banks, settlement infrastructure - these companies are capturing the largest share of the institutional wave. By Crypto Twitter's standards, they may seem boring, but they will generate the most lasting returns for the next decade. The moat of these enterprises is not technology, but regulation. This is something that native crypto players have been underestimating. In the old encrypted world, moats came from liquidity, network effects, and communities. In the new encrypted world, the deepest moat is a license plate. Every jurisdiction that has finalized stablecoin rules, tokenization frameworks, or digital asset banking regulations will create a window - typically 12 to 18 months - in which the first batch of licensed operators establish advantages that newcomers cannot replicate no matter how much money they spend. Customer relationships, banking partners, compliance infrastructure, regulatory trust - these things cannot be forked, and your agent cannot write them. When competitors obtain licenses in the same jurisdiction, the first mover has already locked in the distribution channel. This incident occurred in each jurisdiction, not globally unified. The emerging stablecoin frameworks in Europe's MiCA, Singapore and the United Arab Emirates, South Korea's Digital Asset Basic Law, and the regulatory framework now taking shape in Washington - each creating different license moats in different markets. The companies that can win are those that treat regulation as a strategic asset rather than an obstacle. They recruit pre regulatory officials, not just engineers. They are shaping the framework, not just compliance. Policy proximity - the ability to influence rules in the process of rule making - is currently the most valuable and least understood competitive advantage in the cryptocurrency industry. The native token hype tactics - issuing tokens, attracting TVLs, relying on storytelling to pull shares, and cashing out through unlocking - are coming to an end because the capital entering the system does not play this trick. Institutional allocators require revenue, compliance, and predictable income. They don't want governance over a protocol that may be forked next quarter. As real cash flows move onto the chain, tokens inevitably become like equity. When a protocol generates real revenue and distributes fees to token holders, the token is no longer a speculative tool, but a machine-readable proof of ownership for real business. This is fusion. It's not tokens replacing equity, nor equity replacing tokens, but rather the two collapsing into one thing: programmable, composable, instantly settled, and a statement of rights to real economic activities. The shell is no longer important, what is important is that the underlying business generates cash, and the rights statement for it is software readable. The same license moat also applies to tokenization platforms. When stocks, bonds, and structured products are moved onto the chain, the platform tokenizing them will not be an unlicensed agreement, but a licensed securities intermediary operating under a specific regulatory framework in a particular jurisdiction. The infrastructure is encrypted native, the business model is traditional financial level, and the moat is still a license, not code. Encryption took ten years to build a track for transferring value. But the track for transferring identity, professional competence, and authorization has not yet been established. This is the gap - and where the next wave of infrastructure will be built. The team responsible for authenticating identity, machine verifiable compliance, and portable professional qualifications at the on chain primitive level is building the layer that connects institutional finance and autonomous AI. Both sides cannot run well without it. The AI shift: building becomes cheaper, verification becomes more valuable. The second force is AI, which has a much deeper impact on encryption than implied by the "AI x encryption" narrative. Let's start with the obvious: AI has brought the cost of building any software to an extremely low level. Launching an L2, deploying a set of smart contracts, and launching a DeFi basic module - all of these can be completed in just a few days, and the engineering team only needs a small amount from the past. The impact on existing infrastructure is brutal: when supply approaches infinity, premiums evaporate. More than 100 blockchain systems that have already been launched will be compressed to a utility level profit margin. The infrastructure that used to easily cost $1 billion to $5 billion in FDVs will be repriced as the actual money it earns. The script of "investing in infrastructure, eating narrative premiums, and returning to individual investors" by venture capitalists is structurally broken. But AI has also done something completely different, which is where encryption has gone from optional to necessary. When AI agents can generate infinite transactions, content, identities, and interactions, the cost of forging anything approaches zero. Garbage information and effective signals become indistinguishable, and robot activities and human activities become indistinguishable. In a world filled with infinite machine generated noise, the only way to establish trust is through cryptographic proofs. Zero knowledge technology has transformed from a niche scaling solution to a necessary infrastructure. Privacy protection credentials have evolved from academic research to an authentication layer for every AI involved system. If you cannot prove who you are, that your transactions are authorized, and that your agent is qualified without exposing underlying data, you cannot participate. This is currently the most underestimated proposition in the field of encryption: ZK and privacy technology are not privacy concepts, they are the trust layer of the AI economy. Two forces converge: where tokens, as the operational layer of machines, and AI converge, I believe the deepest opportunity lies. AI agents have started autonomous trading. Coinbase has just launched a wallet specifically designed for AI agents. The x402 protocol enables machine to machine payment. Autonomous systems begin to hold assets, execute transactions, pay for computing power, and interact with financial services without human intervention. These agents require three things to operate. identity Not a username - it is a cryptographic credential that the counterparty can verify within milliseconds without needing to see the underlying data. Who does this agent represent? In which jurisdiction is it operating? What is authorized to do? If you cannot answer these questions with programmable proofs, you can only go back to centralized databases and manual review. When millions of agents are trading simultaneously, this approach cannot be extended. Programmable assets. Stablecoins prove that currencies can be programmed and settled instantly. The same logic extends to treasury bond, stocks, credit and structured products. The agent does not care whether it holds USDC or token treasury bond. What it cares about is that the terms are readable, the rules are programmable, and the settlement is deterministic. Voucher. Today, compliance exists in human judgment and legal documents. It needs to encode regulatory codes as machine verifiable proofs - KYC status, license, jurisdiction authority, risk limits. The only way to deliver these on a large scale is through cryptographic proofs, not centralized APIs. This is where the two transformations I have been describing truly collide. Institutions push financial assets onto the blockchain because programmable infrastructure reduces settlement risks and operational friction. AI pushes economic activities towards autonomous execution. When these two forces meet, the financial object itself must become software. Machines do not buy tokens for speculation. They consume tokens to operate. This creates a completely different demand curve from the history of encryption. Retail speculation is cyclical, and narrative driven capital will rotate. The scale of machine consumption is linked to autonomous economic activities. As AI systems automate more decision-making, trading, procurement, and coordination, the demand for machine-readable financial objects will expand synchronously. What does this mean for our investment? I am not writing this article for abstract analysis. This is the direction in which my new fund is actively investing. The old encryption game was: find the narrative, grab the token, and exit before unlocking. The new game is to find the licensed infrastructure layer that captures stablecoin traffic, build machine-readable primitives required for agent transactions, and establish the jurisdiction where the investment regulatory framework takes shape first. The company that wins the next era looks completely different from the project that defined the previous era. They have licenses, not just liquidity. Having income, not just TVL. There are regulatory moats, not just network effects. By the standards of those who started with memecoin, they are quite boring. By the standards of those who allocate institutional capital, they are opportunities for a generation. The era of token hype marked the beginning of encryption. Institutions and AI will give it the future. The transition between two eras is happening at this moment, and faster than most participants realize. The speed of this transformation is a story that no one has fully told yet. Consider this as the first draft.