CZ's 2025 Year-End AMA: Four Core Focus Areas and His Views on Stablecoins and Prediction Markets

In this end-of-year AMA, CZ reflects on 2025 as a turning point both personally and for the broader crypto industry. He discusses life after his pardon, how his focus has shifted toward education, early-stage investing, ecosystem mentoring, and advising governments on crypto policy. The conversation covers key milestones across BNB Chain, Giggle Academy, YZi Labs, and Binance, highlighting renewed growth on BNB Chain, rising adoption of stablecoins, and the emergence of prediction markets as a meaningful on-chain use case. CZ also shares his long-term philosophy on building–emphasizing mission-driven founders, competition as a net positive, and patience over hype–while offering views on stablecoin evolution, AI agents, real-world asset tokenization, and where crypto infrastructure may head in 2026 and beyond.

The audio transcription is done by GPT and may contain errors. Please listen to the complete podcast:

YouTube：

Spotify：

Host Interview with CZ

What does 2025 mean to you after the pardon?

CZ: The past two years have really felt like a roller coaster. Being pardoned brought a huge sense of psychological relief. Whether you call it freedom, closure, or finally “turning the page,” at its core it’s the feeling of being able to move forward unburdened again. That said, on a practical level, my lifestyle hasn’t changed that much.

Right now, I mainly focus my time on four things.

The first is Giggle Academy, a project I’m deeply invested in and genuinely enjoy. More than 90,000 kids are already using it, and the team is about 60 people. We’re constantly iterating on the product, improving features, and gradually building a community. It’s a completely free education platform and something I plan to commit to for the long term.

The second is YZi Labs. Here, I’m more of a mentor and coach-working with founders and developers, helping them grow. I really enjoy interacting with the next generation of builders. Over time, I’ve also realized that many things I take for granted aren’t obvious at all to first-time entrepreneurs, which actually makes this work even more meaningful.

The third area is the BNB Chain ecosystem, which overlaps significantly with YZi Labs, since many of the projects we invest in are built on BNB Chain. I’m no longer involved in hands-on technical work. Instead, I spend more time talking with teams and founders, helping them calibrate direction, mindset, and long-term goals-especially around mission and values.

The fourth is advising governments, and this takes up far more time than I initially expected. I’ve had fairly in-depth discussions with governments in more than a dozen countries-some formal, some informal; some public, others private. Their main concerns are regulatory frameworks, how to attract the crypto industry, and what they should or shouldn’t do. Crypto is evolving quickly, and there’s still a lot I don’t know myself, but whenever people ask, I’m happy to share what I’ve learned. Pakistan and the UAE are typical examples. Since I live in the UAE, many of those conversations are informal.

These four things take up most of my time. I’ve always worked from home, and that hasn’t changed. Overall, life is in a good place. I also make a conscious effort to stay physically active-that’s a habit I carried forward from that period. I work out almost every day for 30 to 45 minutes and try to do more outdoor activities as well. Without a so-called “formal job,” my schedule is actually more stable now. All in all, life is pretty good.

Looking back on 2025, what milestones are you most proud of-and why are they more than just good-looking numbers?

CZ: If we start with headline metrics, the first one is Giggle Academy. It now reaches roughly 90,000 kids. For a relatively small team, that’s a huge achievement-arguably an impact that already exceeds that of a traditional university.

BNB Chain also saw very strong growth this year. To be honest, it had been somewhat undervalued and overlooked over the past few years. This year, more effort was put back into it, and the results are clear: on-chain transaction volume grew by about 600% year over year-roughly six times. Daily active addresses are now over 2 million; previously published figures were around 2.4 million. Even after discounting for real users, it’s still close to 2 million, which is an incredible scale. BNB Chain’s transaction volume is now among the top in the entire industry.

I don’t spend much time constantly comparing it with other public chains. I generally encourage builders to focus on doing their work well. That said, I do follow BNB Chain’s official channels, so I naturally see these growth numbers.

YZi Labs also went through an important transition this year. Ella moved back from an operator role to focusing on investments, and the overall pace picked up significantly. In 2025, we completed close to 70 investments, evaluated over 1,000 projects, and continued to run the Easy Residency program. Across two cohorts, we supported roughly 30 to 40 projects, all of which are making solid progress.

Binance, even without my direct involvement, has continued to perform very well. Its user base is now approaching 300 million, which is a major milestone for any crypto exchange-and growth is still ongoing.

From a broader perspective, overall crypto adoption remains very low. Whether measured by number of users or by share of global wealth, penetration is still largely in the single digits-and at the global wealth level, it may even be below 1%. That means the industry still has room to grow by orders of magnitude. These are the numbers and milestones that stood out most to me when reflecting on 2025.

Why is now a good time to invest in prediction markets? How do you see the space evolving in 2026 and beyond?

CZ: Historically, prediction markets have always had massive potential-the ceiling is very high. But for a long time, they never truly broke out within the crypto space.

That’s started to change recently with projects like Kalshi and Polymarket, especially Polymarket’s performance during the last election cycle. That was when the category really entered the mainstream. In many cases, its outcomes were even more accurate than traditional polling, because participants are putting real money on the line. That significantly improves the signal-to-noise ratio of the information. So I think this is a very meaningful direction. That said, it’s still hard to predict how things will ultimately play out. One project might emerge as a clear leader, or multiple projects could grow in parallel.

As for why YZi Labs supports multiple projects in this space, the reason is simple: we don’t make exclusive bets, especially at the early incubation stage. We prefer to back strong teams. Teams can collaborate or compete-both are completely normal. What we focus on are fundamentals: the team itself, the problem they’re trying to solve, whether there’s real market demand, and their ability to execute. We also tend to favor founders with a strong sense of mission who are committed to building for the long term.

This isn’t a “winner-takes-all race.” In any market, multiple players usually coexist. Even in centralized exchanges, Binance is large, but there are still many other exchanges. Markets only converge to a single dominant winner after a very long time-if they ever do. At an early stage, it’s actually healthy to let more teams experiment with different approaches. Over time, some teams will merge, some will drop out, and some will break out.

At this stage, supporting more players is really about expanding the overall market and accelerating the industry’s growth. Competition doesn’t usually shrink a market-it often makes it bigger. Different teams enter from different angles, focus on different regions, and attract different types of users. That diversity is especially important for a brand-new category. As for whether to invest in one project or many, I don’t try to force a specific outcome. YZi Labs’ principle is simply to back strong teams.

From the perspective of BNB Chain, it’s a fully open public blockchain where anyone can deploy contracts. As a BNB holder, I naturally welcome more projects choosing to launch on BNB Chain. The BNB Chain team will also provide support in terms of exposure, resources, potential funding, and user growth. If a project is good enough, the ecosystem will support it. Competition is inevitable, but overall this is a healthy and positive dynamic-especially in an ecosystem that’s still in its early stages.

What role does BNB Chain play in the stablecoin ecosystem? And what needs to change when stablecoins are used by AI agents?

CZ: I think stablecoins represent a massive market. What we’re seeing today is still mostly “stablecoin 1.0,” and true 2.0 is only just beginning. The traditional stablecoin model is very simple: you deposit U.S. dollars in a bank and issue an equivalent amount of on-chain tokens. Because they’re blockchain-native assets, they move extremely fast and enable arbitrage across exchanges and across borders-something that’s far slower and more expensive through the traditional banking system.

Stablecoins have already become a huge business, but most mainstream stablecoins today don’t offer any yield. Now we’re starting to see what I’d call “stablecoin 1.5,” where yield mechanisms are built into the design. Put simply, users want stablecoins that are easy to transact with and can also generate reasonable returns. At the same time, stablecoins have very strong network effects. One of the key reasons USDT has stayed in the lead for so long is its first-mover advantage.

However, USDT isn’t particularly competitive when it comes to yield, and that creates an opening for other stablecoins. We’ve already seen a lot of innovation here-Ethena is a good example, and YZi Labs has invested in it as well. Of course, its growth is still constrained by the level of exchange support it gets. In my view, an ideal stablecoin needs to meet three criteria at the same time: it has to be easy to trade, widely adopted by exchanges, and able to offer yield. That sounds simple, but it’s very hard to execute in practice, which is why we encourage more teams to experiment.

Regulation has also changed significantly compared to a year ago. The U.S. now has the GENIUS Act, and many other countries have become much more friendly toward stablecoins. Overall, the regulatory environment is more open, which creates better conditions for innovation.

When it comes specifically to BNB Chain, for a long time it lacked a truly native stablecoin with deep liquidity. USDT on BNB Chain is essentially a wrapped version. USD1 is the first solution that’s genuinely native in a meaningful sense. Before that, there was FUSD, but it didn’t gain much traction due to friction in issuance and distribution. USD1, being backed by the U.S. system, has certain advantages.

More recently, the $U project has also made new progress, and I think it has some potential as well. From BNB Chain’s perspective, it’s an open public blockchain-anyone can deploy a contract and launch a project. The ecosystem’s stance is very clear: it welcomes more stablecoin projects and will provide support where possible. This isn’t a “horse race,” but more like an open garden, where multiple projects can grow at the same time.

What do you value most when supporting new projects? What kind of support can BNB Chain offer to builders?

CZ: At a fundamental level, I prefer to support builders who are genuinely driven by a sense of mission. They need to be passionate about what they’re building. Market narratives will always shift-yesterday it was memecoins, today it might be traditional finance, tomorrow stablecoins. That’s normal. Adjusting the product to find product–market fit makes sense. But if a team completely pivots its direction every three months, it often means they don’t truly care about what they’re building.

Motivation matters a lot. Some people start companies purely to make a quick exit-setting targets like $10 million, $50 million, or even $100 million. The problem is, once they hit that number, what comes next? Many will cash out, walk away, or simply abandon the project. That mindset also makes people more likely to take shortcuts. In crypto, shortcuts are especially tempting-like launching a token and dumping it immediately. These aren’t models we respect or want to support.

What I really look for are founders who are long-term oriented, grounded in strong values, and capable of execution. Building a company is extremely hard. It requires more than intelligence and skills-it requires endurance. In many ways, it’s almost physical labor: long hours, late nights, and sustained exposure to intense mental and physical pressure.

Beyond that, I don’t impose many additional constraints. If a founder has a sense of mission, can execute, and is willing to commit for the long term, that already covers the core criteria I care about most.

Community Q&A

How can someone with a technical background transition into business and leadership roles?

CZ: For people with technical backgrounds, emotional intelligence (EQ) is critical if you want to go far in the business world. You need good judgment about your environment, and you need to know how to communicate and build relationships with very different kinds of people.

Before that, though, I think you have to become truly strong at the technical side first. Everyone needs at least one very solid, well-defined core skill before expanding into other areas. Otherwise, it’s easy to end up knowing a little bit about everything but not excelling at anything, which is usually not an ideal position.

From my own experience, I spent several years at Bloomberg and was involved with multiple startups at different stages. Over time, I gradually built a network of highly capable people across many fields. That didn’t happen overnight-it was the result of long-term accumulation.

There are a small number of founders who catch a break early and scale very quickly, but those cases are rare. For most people, real breakthroughs come only after years of building experience, skills, and relationships.

There’s no standard path for entrepreneurship or career development, and no single “correct” answer. Don’t get overly fixated on any one piece of advice. Opportunities are always out there-keep learning, keep building, and just keep moving forward.

How can BNB Chain move from being “undervalued” to achieving long-term growth?

CZ: Almost every project that eventually succeeds goes through a period of skepticism early on. Whether it’s a public blockchain, a token, or a company, nothing starts at the top. The projects that truly break through are usually the result of long-term accumulation.

Success takes time. Take Nvidia as an example-it took decades to get to where it is today. When you’re building an ecosystem in its early stages, it’s often hard to see short-term returns, and it’s easy to be underestimated.

For BNB Chain, the core strategy has always been very clear: build for the long term and serve real builders. Most of the time, that means staying focused on execution, with occasional course corrections, rather than obsessing over short-term rankings or hype. You can’t always be the most profitable or the hottest project. Mistakes and setbacks are part of the process-the key is whether you can keep moving forward. From day one, BNB Chain was designed as a long-term project.

The ecosystem should be bigger than any single business or application. What really matters is finding the right people and teams and enabling them to keep building on a public blockchain. Through sustained, long-term investment, value compounds over time. Being a builder requires a long-term mindset. It’s less like a sprint and more like a “marathon mixed with a boxing match”-the only way through is to keep building and keep moving forward.

Why isn’t stablecoin competition a zero-sum “horse race”?

CZ: It only looks like a horse race if you choose to frame it that way. In my view, that’s not an accurate analogy. In an open ecosystem, multiple projects naturally coexist-not because it’s deliberately designed that way, but because that’s how markets work. In a space like stablecoins, once people see the size of the market and the profit potential, it’s only natural that many teams enter. The same thing happened with exchanges, search engines, and e-commerce platforms. Competition itself is normal.

BNB Chain is an open public blockchain, while YZi Labs is an investment firm-those are two very different roles. At the early investment stage, it’s almost impossible to back only a single project. So this isn’t a so-called “horse race strategy,” but simply standard investment and market behavior. We support multiple strong teams. Competition is inevitable, and in the early stages it’s healthy-it helps expand the overall market faster and more quickly validates which products and models users actually want.

In the end, network effects will naturally guide users toward the most competitive stablecoins. We’re still in the stablecoin 1.0 phase today, and stablecoin 2.0 will be significantly better.

Will crypto-native social platforms replace X?

CZ: As products mature, functional convergence across platforms is inevitable. Social features are very important for improving user retention and engagement. When users can discuss trades and share views directly on a platform, it often leads to higher participation and trading activity.

For crypto-native platforms like Binance Square, developing social features is both natural and valuable. From a technical standpoint, building chat functionality isn’t hard. But doing it at massive scale-while keeping systems secure, stable, and scalable-still leaves a lot of room for improvement, especially when it comes to privacy and encrypted communication.

I think any large platform will eventually need its own messaging system, not just for social interaction, but also for customer support and user collaboration. At the same time, social platforms adding payments and crypto functionality is also a positive development for the industry.

That said, financial platforms and social platforms have very different “DNA.” Financial platforms prioritize risk management, compliance, and security, while social platforms focus on low barriers to entry and high levels of interaction. Once strict KYC is introduced, social platforms often lose a large portion of their users.

For these reasons, I don’t think one side will completely replace the other. A more realistic path is for each to gradually add features on top of their core strengths, becoming complementary rather than mutually disruptive.

What sets Probable apart from other prediction markets? And will it be integrated with Binance Wallet in the future?

CZ: Honestly, it’s still very hard to identify any truly clear differentiation for Probable right now. It’s a BNB Chain–based prediction market project, and it’s still at a very early stage. The current product form and feature set don’t define what the project will ultimately become. Most early-stage projects iterate constantly before they find real product–market fit.

When I look at projects like this, I don’t focus too much on specific features at the moment. What matters more to me is the team’s ability to execute: can they build a solid product around current demand, can they adapt quickly when demand shifts, and do they have the capability to deliver consistently and pivot when necessary? These factors are far more important than any single feature.

In the prediction market space, features will inevitably be learned from and copied across projects. Designs that truly attract users will be quickly replicated, while features that don’t work will naturally be phased out. That’s why it’s still too early to say which model will win.

From a user’s perspective, the more reliable criteria are fairly straightforward: ease of use, capital efficiency, fees, security, and-at least to some extent-returns.

This is still a very young sector. In the short term, major events (like the World Cup) may create temporary advantages, but in the end, what really determines success is long-term execution and whether the team is willing to stay committed over time.

Looking ahead, how do you see BNB Chain’s goals and direction in 2026–2027?

CZ: I stay in touch with the team and offer help where I can, but ultimately, decision-making and execution are fully in their hands. My input is mostly at a high level, rather than telling them “this specific feature will win” or “this is what you must do right now.”

What I usually focus on are questions like: how the team is structured, how to attract and retain users, and what values and principles the project should stand for. It’s more about mindset, organizational structure, and long-term thinking than about specific product details. To be honest, I haven’t had very in-depth discussions with the team about detailed plans in the past few weeks, so I can’t really comment on a concrete roadmap. From what I can see, though, the team’s interaction with the community is relatively open and proactive.

How do you see DePIN, payments, and real-world assets shaping the next phase of crypto-especially after 2026?

CZ: From what I’ve seen so far, real-world assets (RWA) and asset tokenization are among the most seriously pursued directions in the industry, and they’ve moved well beyond the conceptual stage. I’ve personally spoken with senior officials in multiple countries, and tokenizing sovereign or national-level assets is not only feasible in practice, but also makes clear economic sense.

The core value of asset tokenization lies in its ability to enable financing and kick-start development. Many countries hold highly valuable resources-such as rare earths or energy assets-but lack effective ways to monetize them or build industries around them. Through tokenization, these assets can attract early capital and attention, which can then be used to fund industrial development and technological progress. The key condition, of course, is that the capital must be used responsibly-otherwise, it’s simply borrowing against future resources.

By comparison, I’m less directly involved in DePIN myself, but it has a strong connection to AI and robotics. AI and robotics are areas where we’re investing heavily, and YZi Labs has made a number of bets in this space. Over the long term, AI agents and robots will inevitably rely heavily on crypto, because they can’t use the traditional financial system-they can’t open bank accounts, go through KYC, or use credit cards.

From that perspective, crypto is a payment and value system that’s naturally designed for autonomous agents, while DePIN serves as the bridge between crypto networks and the physical world-infrastructure, hardware, and robotic systems.

Overall, across DePIN, AI, payments, and other themes, real-world asset tokenization remains the most mature and most practically actionable direction at this stage, and continues to stand out as a core focus for the industry.

Will AI trading agents become a mainstream narrative in 2026?

CZ: AI has already been widely used in trading for a long time. Large trading institutions, quantitative teams, and hedge funds all rely heavily on AI-or AI-like approaches-for data analysis and decision support, even if they don’t always label it explicitly as “AI trading.”

The real uncertainty is whether “shared” or subscription-based AI trading agents can become a mainstream product model. If an AI strategy can consistently make money, the most rational approach is usually to use it internally and trade with it directly, rather than selling it to others-unless the team is severely capital-constrained. But today, if a model is truly strong, raising capital isn’t that difficult.

When AI trading systems are sold to large numbers of users on a subscription basis, it often implies one of two things: either selling the AI is more profitable than trading with it, or the overall returns of the strategy aren’t actually sufficient to cover all users’ fees and expectations.

Some argue that allowing users to customize or personalize AI models can avoid the problem of everyone using the same strategy. I’m personally skeptical of that. Minor parameter tweaks rarely create a durable, long-term trading edge.

Another fundamental issue is that trading strategies can’t be infinitely replicated. A strategy may work well when only a small number of participants use it, but once it’s widely adopted, it quickly breaks down due to market self-interaction. Trading is highly dependent on execution speed, market structure, and participant behavior. When a strategy becomes “crowded,” early users may profit, while later users often lose money.

This is why AI trading is fundamentally different from prediction markets. Prediction markets are platform businesses that earn through fees or spreads, while trading strategies themselves tend to lose effectiveness as more people adopt them.

In short, AI will definitely be deeply embedded in trading, but it’s unlikely to explode in the form of “everyone using the same AI trading agent.” A more realistic outcome is that every professional participant will use AI-but in different ways, with different models and strategies.

Follow us
Twitter: https://twitter.com/WuBlockchain
Telegram: https://t.me/wublockchainenglish