Binance founder Changpeng Zhao, or CZ, used a Sept. 25 appearance on the When Shift Happens Podcast to revisit the path from his family’s immigration to Canada to the creation of Binance, while also addressing FTX, Hyperliquid, HYPE’s absence from Binance Spot, the Oct. 10, 2025 market crash controversy, his U.S. legal case, and why he now spends more time on regulation, investing and education.
The interview stretched across Binance’s early operating culture, why most listed crypto projects fail, whether retail investors should be allowed into early-stage deals, how the number 「4」 became a crypto meme, how he viewed Bybit’s response to a major hack, and why he believes the crypto market is still nowhere near saturation.
From an immigrant household in Canada to 「the crypto guy」
CZ opened by describing a life that still involves constant travel. He said he usually stays in one place for only one or two nights before moving on, though he added that his schedule is easier now than it used to be and that he has more time he can control himself. Decisions on where to go and which events or podcasts to join are usually handled by his team, he said.
Asked what the biggest change in his life might still be, CZ said he believes the biggest change has not happened yet, even if he does not know what it will be. He said future opportunities are always greater than past ones, and that the pace and scale of change keep increasing. Before the internet, it was hard to make the whole world change at once. Now information moves more freely, blockchain has created a network for money, and AI is adding another layer of acceleration through efficiency gains.
When asked who he is, CZ described himself as a 「crypto guy」 and also a 「crypto sales guy.」 He said U.S. President Donald Trump once called him 「the crypto guy.」 Zhao said his biggest contribution so far has been in blockchain and that most of his future effort will likely stay there. He added that he now travels to many countries to discuss crypto regulatory frameworks, which is one reason people often see him as someone promoting crypto adoption.
Family background, money pressure and motivation
CZ pushed back on wealth rankings that have placed him among the richest people in the world. He said he does not think those lists are accurate, adding that he could be in the top 100 or the top 500 and that he does not know. He also said outside estimates of his net worth are materially overstated.
On what drives him, Zhao said he cannot point to any severe childhood trauma, but he did grow up in a family with limited means. After his family immigrated to Canada, both of his parents earned close to minimum wage. Some of the friends around him came from wealthier families from Taiwan, Hong Kong or elsewhere who had immigrated through investment programs, so there was some comparison and some pressure when he was young.
Still, he said he does not see his childhood as especially bad. He said he had many good friends from Taiwan and that their families treated him well. What stood out to him more was the contrast in economic conditions. He said he sees himself as fortunate because he stood on his parents’ shoulders. In his telling, his parents moved to Canada so that he and his sister could have a better future, making him effectively a second-generation immigrant. He said second-generation immigrants often see how hard the first generation had to work, especially when language barriers push them into difficult jobs near minimum wage, and that this can create a strong drive to keep moving forward.
Zhao also spoke about his sister, who he said has a technical background. She worked at a startup in Tokyo, later joined Morgan Stanley, and became a managing director at a relatively young age. He described that as a very successful career path, noting that Morgan Stanley has only a few hundred managing directors globally despite having a very large workforce. He said she retired at around 42, went through postpartum depression after having a child, recovered fully about a year after leaving work, and now helps advise people dealing with depression.
Asked how he plans to keep later generations of his own family from losing the drive that often comes with immigrant hardship, CZ said he does not yet have a real answer. He said he thinks about it from time to time and suspects education is part of the answer, but called it a difficult problem.
How Binance actually started
Zhao said Binance did not begin in the same way as many startup stories. Before launching the exchange, he already had a team of roughly 15 to 20 people working at Bijie Tech, a company that sold exchange systems to other trading venues. He said around 80% to 90% of Bijie Tech’s clients were not crypto exchanges at all, but traditional marketplaces trading things like stamps or jade while still using order-book systems.
He had been leading that team for about two years, and the business was stable. Around May 2017, he called everyone into a meeting room and told them it might be the right time to build a crypto-to-crypto exchange. He said he told the team that if they did not do it then, the idea would keep coming back, and that if he did not do it at that point in his life, he would probably end up doing it later anyway.
In his view, the team already had many of the right ingredients: a product, a system and a team. What it lacked were a marketing team and a customer support team. Bijie Tech was a B2B business, so it had sales staff and account managers, but not a support structure built for large numbers of retail users. Zhao said everyone in the room agreed to do it, and the group then discussed domain names before choosing Binance from two options. That, he said, was the real starting moment for Binance. The room itself was small, possibly smaller than the room where the interview was being recorded.
Asked what allowed Binance to go from zero to the world’s largest exchange in about half a year, CZ said he cannot compare directly with other exchanges because he does not know what their offices were like. He said there was nothing magical about Binance’s setup. What mattered was a team that worked hard, stayed grounded, had little ego, and executed extremely well. He said the team was highly aligned, trusted him, and was willing to follow him through the process.
Zhao also said there may have been an Asian cultural element. In Western settings, he said, decisions often involve extended discussion and broad consensus-building. In Asian settings, once people decide to follow a leader, they may do so very firmly, sometimes so firmly that it becomes hard to get them to challenge the leader’s ideas. If the direction is right, that can create very high execution speed. If the direction is wrong, it can become a disaster. He said Binance was fortunate that it broadly chose the right direction.
On what 「working hard」 means, CZ said hours alone are not the right metric. What matters more is whether someone produces results and does so efficiently. Many people work long hours without building anything meaningful, he said. The real test is whether they can create high-quality products that users genuinely like. He added that passion matters because people who care deeply about what they are building often do put in long hours, but the better measure is the quality and speed of output.
He said that in Binance’s early days, and even now, many people did not count their hours. If something had to be finished, they would put their children to bed at 9 p.m. and then keep working until midnight or 1 a.m. if needed.
Misconceptions about Binance and the role of centralized exchanges
CZ said there are many misunderstandings about Binance. One is the idea that Binance is a Chinese company. He said it is not, even though the company has many Chinese employees. Another is the belief that centralized exchanges are bad for crypto or even inherently evil. He said he does not share that view.
According to Zhao, Binance has always supported decentralization. The reason centralized exchanges remain dominant, he said, is that most crypto users still prefer email, passwords and customer support over managing wallet addresses and private keys themselves. Even so, he said he believes the industry will become more decentralized over time, which is why Binance has also invested actively in decentralized ecosystems.
Why most listed projects disappoint
When asked why so many projects listed on Binance end up disappointing investors, CZ said the answer is not unique to crypto. In any market or industry, most projects fail. He compared crypto with the internet era, when millions of internet companies were created but only a few hundred became meaningful successes and only a few dozen became truly large. He said AI will likely follow the same pattern: most companies will fail, but the winners could become extremely large.
The host pointed out that tokens differ from early internet startups because tokens are public and anyone can buy them, which means anyone can lose money on them. Zhao said that cuts both ways. He asked whether people would have wanted the chance to invest in Anthropic when it was still very small. In his view, most companies in any new industry fail, but the industry itself survives and a small number of firms do very well.
That led to the question of whether ordinary investors should be allowed to invest early. CZ said he does not agree with the logic behind 「accredited investor protection,」 where people are kept out of early-stage opportunities in the name of safety. He said that approach may actually leave many people poorer, though he added that it may not apply equally to everyone.
He pointed to the U.S. stock market as an example. Many ordinary Americans have been able to build wealth in part because they could participate in equities, he said. If that opportunity were removed, many would have less wealth. In many other countries, local stock markets are weaker than the U.S. market, and people often cannot easily access U.S. equities either. So the real question, in his view, is whether society should expand access to investment opportunities or restrict participation in the name of protection.
Zhao said he does not claim to have the one correct answer, but his own instinct is to give more people access to investment opportunities, provide them with enough education, and let them decide for themselves.
Can retail still win when market makers, KOLs, launchpads and exchanges are involved?
CZ said the short answer is yes, retail investors can still win, but he stressed that the issue is full of nuance. He used advertising as an example. Advertising can be done well or badly. If it makes false promises, it harms users. But the line between false and true is not always black and white. Projects can overstate things or use vague language that leads people to the wrong conclusion.
He said healthier markets require several things at once: founders doing the right thing, users being educated on how to evaluate projects instead of chasing hype and marketing, and a regulatory framework that imposes disclosure standards. Traditional stock markets in most countries have disclosure requirements, he said, while crypto markets are still less mature on that front.
Even so, Zhao argued that over the long run, markets tend to filter out projects built mainly on promotion. Some people will always be very good at hype and marketing, and some of them may profit in the short term at users’ expense. But projects that last are the ones that build sustainably. In the end, he said, the market decides.
How 「4」 became a meme
Zhao traced the 「4」 meme back to a 2023 post laying out his New Year priorities. He said he wanted to do fewer things and focus mainly on education, regulation, compliance and product. Item No. 4 in that list was to ignore FUD, including fake news and negative narratives.
He later said that whenever he saw that kind of FUD again, he would just post the number 「4」 and let the community explain it. The community embraced the idea quickly. He said a negative report appeared less than 24 hours later, and people immediately started replying with 「4.」
The meme then spread through community participation. One member posted a selfie holding up four fingers, which Zhao said he thought looked cool and which the community liked. A few weeks later he posted a similar photo himself, and the meme spread further. Two years later, with meme culture much hotter, he said there may now be thousands of meme projects with 「4」 in the name.
Asked whether there had ever been a time when he posted 「4」 while privately wondering if the criticism might contain some truth, CZ said that most of the time he uses it because he knows there is something inaccurate in the report. He added that he does not think he has abused it so far, though others may disagree.
On the claim that one tweet killed FTX
The interview revisited Zhao’s Nov. 6, 2022 tweet saying that as part of Binance’s exit from its FTX equity investment, Binance had received about $2.1 billion equivalent in BUSD and FTT, and that due to recent revelations it had decided to liquidate the remaining FTT on its books. FTX soon collapsed, and Sam Bankman-Fried later said CZ’s tweet killed the company.
Zhao rejected that outright. If a competitor can kill your company with a single tweet, he said, then you never really had a company. A real business cannot be destroyed that way. He said he could tweet about any company and it would not collapse just because of that.
In his view, FTX failed for many reasons. He said he does not know every detail, but no company disappears simply because a rival posts on social media, no matter how large that rival is. He compared it to a hypothetical AI chip company. Even if Jensen Huang tweeted that he had once invested and was now selling the stock or token, Zhao said, a company with a good product, real customers and healthy cash management would not vanish. The price might move in the short term, but the company itself would remain.
Asked whether he had underestimated the impact of his own tweet, CZ said he does not think the tweet had that level of power. He described it as one factor among many. CoinDesk had already published a report three or four days earlier questioning FTX’s solvency, he said. The real issue, in his telling, was that FTX had not told customers the truth and had misused customer funds. Customers were told their money was there, but the funds had been used to buy other things, leaving the company without enough liquidity when withdrawals came.
Zhao said he did not know those details at the time. He also noted that Binance’s exit from the FTX equity investment had not happened 「the year before」 but roughly five years earlier from today’s vantage point, even though Binance still held FTT. Since Binance was preparing to deal with those tokens, he said, he believed it was better to disclose that plan publicly and transparently. That is why he saw the tweet as a disclosure.
He said he did not expect the market reaction that followed. The tweet explicitly said Binance would sell over a period of months to minimize market impact. He added that he never knows how markets will react and did not foresee the volatility that came next. He also noted that Alameda CEO Caroline Ellison posted her own tweet about 20 minutes later, and that some people have argued her post revealed more than his did.
As for what people still misunderstand about the week FTX collapsed, Zhao said he does not know of any hidden internal facts beyond the core issue as he understands it: billions of dollars in customer funds were used to buy other things, and when markets fell and customers wanted to withdraw, the money was not there.
BNB below ICO, the U.S. case and prison
Asked about the hardest day of his life, CZ said there were several difficult periods. One came shortly after Binance launched, when BNB began trading publicly and fell below its ICO price for about two to two and a half weeks. He said that period was psychologically difficult because thousands of people had bought the token and entered the industry only to see immediate losses. He said the pressure was intense, though the period was relatively short.
Another difficult stretch came later, when he was dealing with the U.S. government. He said that period lasted about a year and a half to two years and carried heavy pressure.
On his time in prison, Zhao said the dominant feeling when he entered was not a belief but a fear: that he might never get out. His biggest concern was that one more charge could be added, then another, and another, until the process became endless. He said that fortunately did not happen.
He argued that, to his knowledge, no one in U.S. history had been sent to prison solely for a single Bank Secrecy Act violation. Even today, he said, many similar cases do not lead to criminal prosecution, and many bank executives are not prosecuted. Some may receive deferred prosecution agreements instead. He cited Arthur Hayes as a relatively close comparison, noting that Hayes, another 「crypto guy,」 received home confinement.
Zhao said that because Hayes’ case unfolded during the pandemic, it may have been easier in some ways. He had thought that kind of outcome represented the harsh end of the range, but his own case turned out differently. Although his sentence was four months, he said that did not fully remove the uncertainty because he could not know whether another agency might bring another charge. Some parts of the U.S. government gave him assurances, he said, but one agency’s promise does not bind every other agency in such a large system.
Asked whether people disappeared from his life when things got hard, CZ said he did not see that among his own friends, and not inside Binance either. He said the earliest Binance co-founders from 2017 are still there and that the team remains very tight-knit. Across the broader crypto industry, however, he said many people did disappear after making grand promises.
How the presidential pardon process unfolded
The host said U.S. presidents sign a certain number of pardons each year and noted that Zhao later received a pardon from President Trump. CZ first challenged the framing, saying the numbers cited were off. Based on what he had read, he said President Biden granted pardons or commutations to a very large number of people, around 1,700 by his recollection, far more than 40 per year. He said he did not know how many Trump had granted.
As for his own case, Zhao said he first discussed the process with his lawyers. He pointed to Trump’s campaign pledge to pardon Ross Ulbricht, the founder of Silk Road, whose case was much more serious and involved a platform tied to drugs and weapons and a life sentence before the pardon. Arthur Hayes was later pardoned as well, Zhao said, and Hayes’ case was much closer to his own because it also involved the Bank Secrecy Act.
According to information Zhao said he read online, Hayes may have been more directly involved in certain transaction-related conduct, though CZ said he was not sure whether every detail of that description was accurate. He said Hayes and several co-founders were pardoned, and that Hayes received his pardon in March 2025.
Zhao said he had already been discussing the issue with his own lawyers before that. A friend later recommended a lawyer who had handled pardon matters for both Hayes and Ulbricht, and Zhao contacted the same lawyer. Because some conversations are protected by attorney-client privilege, he said he did not want to disclose them. But he described the broad process as preparing and submitting a formal pardon application.
He said he submitted his application in April 2025, after Hayes had been pardoned. From there, it was mostly waiting. The lawyer would update him every two weeks or every month, he said, usually just to say that they were still waiting. No one knew when a result would come.
Asked whether he had to meet Trump personally, Zhao said no. He said he has never met Trump, never spoken with him by phone, never emailed or texted him, and does not even have his phone number. The closest he ever came was at Davos, where Trump was on stage and Zhao was in the audience. They did not shake hands or exchange greetings, he said, and Trump would not have known he was there.
When the host suggested some people see the pardon as proof that political influence matters more than justice, CZ said his own view is obviously biased and that others should judge for themselves. He said many people have told him his treatment was unfair. In his account, no one had previously gone to prison solely for a single Bank Secrecy Act violation.
He said the core allegation against him was that Binance had not built a strong enough KYC and anti-money laundering system. He stressed that he did not personally process transactions and that all trading was handled by systems. The government’s position, as he described it, was that the system’s AML controls were not strong enough.
Zhao then contrasted that with the current market. Based on the information he has today, he said Polymarket’s international product does not use KYC, while its U.S. version does, and that the international product relies more on geofencing. He said similar discussions now exist around international perp DEXes, where KYC may not be required in the future. Standards change across time, governments and regulatory priorities, he said, and because he has a direct interest in the issue, he again said others should make their own judgment.
The host summarized that as meaning Binance may have run into trouble because it was early and the rules were immature, while regulators now have to figure out how to handle perp DEXes and prediction markets. Zhao agreed. He said Binance faced problems then because its KYC was not strong enough, while some platforms today have no KYC at all. Looking back through today’s lens, he said, what he did might not be viewed the same way. But he added that standards differ by era, government and policy focus.
If X launches payments, CZ would back Elon Musk
The host noted that Zhao had invested about $500 million in Elon Musk’s acquisition of Twitter, now X, and asked what he would tell Musk not to do if X Money launched with Bitcoin at its core. CZ’s answer was simple: 「Go for it.」
He said Binance was a relatively passive investor in that deal and that he would not tell Musk what he can or cannot do. Musk is a very strong entrepreneur, Zhao said, and if X moves in that direction it would be good for the broader industry and for crypto.
Asked how far away that might be, CZ said he does not think it is far off. He said he has already seen discussion on X about accepting stablecoins, and once stablecoins are accepted, adding Bitcoin, Ether, BNB or other assets may be only one step away from a technical standpoint.
Bybit, hacks and Binance’s 「big red button」
The interview also returned to Bybit’s roughly $1.5 billion hack. The host said CZ had publicly suggested that Bybit pause withdrawals after the incident, while co-founder Ben Zhou chose not to do so. Looking back, the host asked whether that meant Ben had been right and CZ had been wrong.
Zhao said there is no absolute right or wrong in that kind of situation. Ben had more internal information and was making decisions with his technical team, he said. Since the situation did not worsen, keeping withdrawals open did not create additional problems. But Zhao said his own style is more conservative. He would pause first and investigate after.
That led to Binance’s own 2019 hack, when about $40 million was stolen in May of that year. CZ said the first thing he did then was pause all withdrawals and, in practice, stop the whole system. In his view, every exchange should have a virtual 「big red button.」 When a security incident happens, someone should be able to hit that button and stop everything immediately, then investigate what happened and how far the damage goes, rather than leaving the system running while trying to understand the breach.
He said anyone responsible for the relevant systems inside an exchange should have the authority to do that. Binance had such a mechanism at the time, he said, and he believes it still does. Only after the facts are clear and confidence is high should operations be reopened gradually.
Zhao acknowledged that this approach disrupts trading continuity. On a futures platform, prices keep moving, liquidations that should have happened may not happen, and other distortions can appear. But he said those issues are secondary compared with the risk of a live vulnerability causing more funds to leave incorrectly.
He said Binance kept withdrawals paused for about a week in 2019, though trading resumed after a few hours because the exchange did not want its prices to drift too far from the rest of the market. During the first hours and days, the team inspected the system very carefully. Even after they felt they broadly understood the issue, Zhao said he chose the more conservative route and kept withdrawals closed for a week.
During that week, Binance also made major upgrades and even changed parts of its architecture from a more traditional server model to a more serverless one, though Zhao said he did not want to go deep into the technical details. Looking back, he said one could argue Ben Zhou’s decision was technically correct because no second incident followed. But if there had been a second attack, the outcome would have looked very different.
Asked whether the result would have changed much if Bybit had followed his advice, Zhao said not necessarily, if there was only one attack. The big difference would have come if a second attack had happened. Fortunately, it did not. He added that Bybit was using a third-party wallet provider and that, as he understood it, the code related to the attack may have been in place for some time rather than inserted on the same day. The attacker may have altered part of the UX or code in advance and waited for the right moment to trigger the exploit.
Zhao said his public call for a withdrawal pause was also meant to help Ben. If Ben had announced a pause on his own, he would have faced heavy user pressure. If someone with Zhao’s profile publicly said a pause might be the safer option, that could give the decision more legitimacy and reduce complaints. He described his suggestion as a more conservative path that would inconvenience users but improve safety.
Hyperliquid is not the enemy, CZ says
On the claim that Hyperliquid is the first real competitor Binance has faced, and that some of his public comments were meant to slow Hyperliquid down, Zhao said flatly that this is wrong. He said he wants more exchanges and more trading platforms to appear and innovate.
His reasoning is that the crypto market is still far from saturated. By headcount, he estimated that perhaps 5% to 15% of people own some form of crypto asset. By share of net worth, he said, crypto likely accounts for less than 1% of most people’s wealth, perhaps only a fraction of 1%. That means the market is still early, and the industry should want as many participants as possible helping to expand the pie.
Zhao acknowledged that some people in the Hyperliquid community frame centralized exchanges, and Binance in particular, as evil in order to attract users who dislike Binance. He said that is simply one way to compete. But he insisted Binance is not against Hyperliquid. When Trump mentioned Hyperliquid, Zhao said, he publicly posted that it was great because it was good for the whole industry.
He said that when HYPE rises, other assets often rise too. BNB rises, Bitcoin rises, and overall activity across the sector picks up. He said he hopes that kind of momentum can become a turning point that helps the market emerge from a slump.
Zhao also repeated a broader point he has made before: the first mover in a category does not always become the long-term winner. First-generation products do a lot of the hard work and make a major contribution, but later generations are often better optimized. Google was not the first search engine, Facebook was not the first social network, and Binance was not the first centralized exchange. Even in DEXs, he said, Uniswap now faces rivals that can match it in trading volume.
For that reason, he said, he is not Hyperliquid’s enemy and would actually like to see it keep opening the path. If what it is doing in the U.S. pushes the industry forward, he said, that is a good thing. Even if some people in its community attack Binance, Binance is not treating them as enemies. More perp DEXes and more competition are good for the industry, he said, and the bigger picture matters more than the growth of any single platform.
Asked what Hyperliquid does better than Binance today, Zhao gave a short answer: 「They don’t have KYC.」
Why HYPE is still not on Binance Spot
The host noted that Hyperliquid founder Jeff had appeared on the podcast around the time HYPE launched and that HYPE still has not been listed on Binance Spot despite being, in the host’s words, one of the best-performing assets of the past two years, up about 25x and now among the larger tokens by market value.
Zhao said he had already left Binance’s management role by the time HYPE appeared, so he does not know the exact internal reasoning. But based on his understanding, HYPE was not fully freely circulating for a long period and existed mainly within Hyperliquid’s own chain and infrastructure. Early centralized exchanges that listed it may have had to keep HYPE within that infrastructure.
Binance’s policy, as Zhao described it, may require the exchange to be able to directly control and hold the assets it lists. The point is to reduce counterparty risk. If a project or its infrastructure disappears one day, Binance would still want to actually hold the tokens. He said he believes Binance has long had that policy. Whether that is still an issue for HYPE today, he said, he does not know.
The Oct. 10, 2025 crash and why Binance became the villain
The host then turned to the market crash of Oct. 10, 2025, saying many altcoins fell close to zero within minutes and that the event produced one of the largest liquidation waves in crypto history. As in other financial crises, the host said, the market seemed to want a villain, and after 10/10 many people chose Binance.
Zhao said the reason Binance was singled out is fairly obvious to him: a competitor was aggressively pushing the narrative that Binance caused the problem. But if one looks at the price action, he said, the initial move started elsewhere.
He laid out the sequence as he sees it. First, President Trump announced tariffs. Stocks fell, and crypto fell with them. Then Binance had a relatively small technical issue involving a stablecoin. The host identified that stablecoin as Ethena’s USDe, and Zhao agreed. He added that USDe was not on the scale of USDT or USDC and was much smaller, in the tens of billions of dollars range.
Zhao said Binance did have some issue related to that asset, but that it was not the cause of the broader 10/10 crash. Media coverage then began attributing everything to Binance, he said. Cathie Wood mentioned in a podcast or interview that Binance may have been one factor. Zhao said he thinks she was probably reacting to news reports rather than a deep review of the timeline. He added that Chinese-language media and a Chinese competitor operating a centralized exchange amplified that line heavily.
He said he later recorded another podcast with Cathie Wood, and that in that conversation she stated clearly that she did not believe Binance caused the entire market crash. Rather, she said Binance had a technical issue that formed one component of the broader event.
Zhao also pointed to a chart posted publicly by Dragonfly’s Haseeb. According to that timeline, prices had already started falling after Trump’s tariff announcement, while Binance’s issue came later. So Binance may have had a localized problem, he said, but it was not the starting point of the market-wide sell-off.
What happened next, in his view, is that people wanted a villain, especially after losing money. He said Binance later put up about $800 million to compensate some users on the platform who suffered losses. Binance had no obligation to do that, he said, but did it anyway. That part was easy to overlook because people prefer negative narratives.
Zhao argued that actual Binance users understood what happened. Social media narratives can be bought and influenced, he said, but the real thing to watch is capital flow. During the period when Binance was under heavy criticism, the exchange still saw net inflows, users continued depositing funds, and Binance’s market share did not fall. In his view, that shows the people actually using Binance did not believe Binance caused the event.
Asked whether he posted 「4」 at the time, CZ said no. He said he usually uses 「4」 when a major traditional media outlet publishes a clearly inaccurate negative report. If the criticism is coming from many smaller voices, community accounts or KOLs, he usually does not bother.
He added that many KOLs had sent him screenshots showing offers of $20,000 or $50,000 to speak negatively about Binance for a month. He said some KOLs have publicly said they received such offers. Some may have taken the money and posted negative content, he said, but that level of activity still would not usually prompt him to post 「4.」
What comes next: regulation, YZi Labs and Giggle Academy
Asked what contribution he would want to leave behind if Binance disappeared tomorrow, Zhao said he does not define it in one fixed way. He simply tries to do what he can. In practice, he said, even with Binance still operating, he now spends almost no time on the centralized exchange itself.
Instead, he said, much more of his time goes into talking with governments in different countries about crypto regulatory frameworks, which he sees as helpful to the whole industry. He also spends a lot of time investing with YZi Labs, with roughly 70% to 80% of those investments still in Web3, and he talks with many founders. He also spends time on Giggle Academy. Whether Binance exists or not would not change his next steps very much, he said.
When asked what he most wants to optimize in life now, Zhao answered: positive impact. He said making a positive difference for other people creates happiness in itself. If he knows he is contributing and helping others, he feels happy internally. So the question he asks is what he can do to maximize that positive impact.
He said he does not have a specific number or a target he must hit. He just wants to do his best. If he has strong capabilities, he should do more. If his capabilities are limited, he should do less. But as long as he knows he has tried his best to make a positive contribution, he is satisfied.
In the short term, Zhao said the highest-impact work is helping countries build more reasonable crypto regulatory frameworks that are friendlier to innovation while still protecting consumers. In the long term, he said Giggle Academy may have the greatest positive effect because it aims to provide free education to everyone in the world and is growing very quickly.
He said the project has only been running for about a year and a half to nearly two years. At the start of 2026, Giggle Academy had about 100,000 child users. It now has about 1.3 million, he said. That means growth has been very fast in less than seven months and close to exponential. Over a 10- or 20-year horizon, he said, that could create a very large positive impact.
Why CZ thinks AI can rebuild parts of education
Zhao said he holds some views on education that may sound counterintuitive or controversial. In his view, today’s education system is built to produce an 「average」 and predictable person, in a way that resembles the industrial era’s push to create standardized workers. He said he is not looking down on workers and respects any legal way of earning money, but he believes the future world will be highly competitive.
He gave the example of software developers. A developer may be competing with 10 million developers worldwide, he said, and in that environment it is better to be in the top 1% or even the top 0.1%. At the very top, even a small edge can produce a huge difference in rewards. Moving from the top 0.2% to the top 0.1% can mean a multiple difference in income or opportunity.
That is why specialization matters, he said. Yet schools often do the opposite. If a student is very strong in math but weaker in English, schools often ask that student to spend less time on math and more time catching up in English, pulling everything back toward the mean. Zhao said people still need basic language ability and a baseline level in other subjects to function in life and work, but they should also become highly specialized in one or a small number of things they are truly good at.
His second point was about the teacher-centered model. He said it is expensive and, on average, not very effective. There are excellent teachers, he said, but average teacher quality is not high enough, and one reason is pay. Many of the smartest and most capable people choose entrepreneurship, finance, technology or senior banking roles instead of teaching.
Based on the data he has seen, Zhao said average teacher pay in the U.S. is around $30,000 a year, which he described as already relatively high by global standards, with many other countries paying less. As a result, society is not drawing its most capable people into educating the next generation at scale. The system is also inefficient, he said, because one teacher may face 30 or even 50 children who all learn at different speeds and levels.
For those reasons, Zhao said he strongly believes in AI-driven, app-driven education that adjusts content to each child and delivers it through digital platforms. If done well, he said, it could be much more effective than the current system.
Asked whether he imagines a digital platform or a future robot spending two or three hours a day with a child and helping that child grow one-on-one, Zhao said the key is personalization. In theory, both robots and apps can understand a child’s current state and adapt lessons accordingly.
He said robots have some advantages. In the future they might accompany children in physical education and support more dynamic face-to-face teaching. Apps have weaknesses, especially around social interaction, but he said some of that can be reduced through app-based social features, by organizing children and parents together, and by using digital tools to coordinate offline sports classes or other activities.
Zhao said robots are definitely part of the future, but current robotics still feels clumsy and not yet natural enough. He said children do like robots, but once they realize 「this is a robot,」 they do not interact with it in the same way they would with a real person. By contrast, children are already very comfortable interacting with apps.
He said the same pattern appears in content. AI can already generate children’s storybooks with a click, producing a storyline, images and a 20-page book. But he said children do not seem to like stories that are fully generated by AI. They respond better when a human carefully designs the story and AI is used to generate images or speed up parts of the process.
That is why, for now, Zhao sees apps as easier to scale. Once an app is built, he said, the marginal cost of serving 100,000 users versus 1.3 million users is almost the same. Software scales far better, and building apps for hundreds of millions of users is something he knows how to do. Robots involve hardware and many other issues he is not ready to handle yet. But once robots have natural facial expressions, natural control and high-fidelity interaction, he said, Giggle Academy would definitely use them.
「I can’t move the whole crypto market with one tweet」
Near the end of the interview, the host noted that about 25% of traders on Kalshi think Bitcoin will break $100,000 before January 2027. He also brought up Zhao’s Aug. 19 post, made when Bitcoin was roughly in the $56,000 to $64,000 range, saying 「If you want your future self in two years to thank your current self,」 after which the market rose about 20% over the following days. The host asked whether CZ can move a roughly $3 trillion crypto market.
Zhao said no. He said he has posted too many bullish tweets for that theory to hold. He described himself as consistently bullish and said almost all of his posts lean that way, sometimes five or even 10 in a day. That particular post, he said, just happened to appear about five hours before Trump gave a very positive speech. He said he had no idea the speech was coming or what it would contain.
If he had advance knowledge of market-moving information, Zhao said, he would not tweet about it. He said some people still think crypto broadly follows a four-year cycle and some even calculate that October would be the bottom. His own thought at the time was simply that there was no reason to wait until October to say something positive.
He said he posted the line while traveling and feeling a little bored. He had been reading a book with a line that roughly meant that if you want your future self to thank your present self, you need to act today. He no longer remembers who said it, but he thought it could be adapted into something more upbeat for the crypto community, so he posted it.
About five hours later, Trump delivered a speech that was very positive for crypto and mentioned Hyperliquid among other things. Zhao said he had no prior knowledge of the content and that no one had told him anything in advance.
His conclusion was simple: if you are always bullish and post five bullish tweets a day, eventually one of them will land five hours before a major positive catalyst. The host closed by saying that at least nobody is complaining about that tweet today, and Zhao agreed that people are happy now.