Summary

Binance’s founder story is less lightning in a bottle than a long accumulation of immigrant frugality, trading-system engineering, and constrained pivots. CZ went from earning C$4.50 an hour at McDonald’s to optimizing microseconds in Tokyo, managing 60–80 people at Bloomberg, and spending eight years building a Shanghai IT consultancy that could fix printers or implement SAP. The episode’s broader conclusion is that most successful businesses are not instant college or garage breakthroughs.
CZ converted his Bitcoin conviction into a concentrated personal bet before he had a durable crypto business. After six months of study, he sold his Shanghai apartment for roughly $900,000 and bought BTC in tranches from $800 down through $600 and $400, averaging about $600. The logic came from Bobby Lee: risk 10% of net worth with a small chance of losing it, versus a higher chance of a 10x that doubles the whole portfolio—yet CZ initially thought, “I’m too late.”
Binance emerged when regulation erased a profitable exchange-software business and an ICO supplied both capital and distribution. After about 30 exchange clients produced recurring SaaS revenue, a March 2017 Chinese shutdown eliminated most customers; a 20-person team then pivoted to its own crypto-to-crypto exchange. Roughly 20,000 buyers—CZ estimates 80–90% Chinese—purchased 60% of BNB for about $15 million, with a 50% trading-fee discount as the immediate utility.
The early moat combined visibly faster execution, BNB incentives, and a user metric that resisted short-term monetization. Orders were observably faster than on competing platforms, BNB recovered from about $0.06 after pricing near $0.10, and revenue quickly reached hundreds of BTC. CZ’s operating north star was “daily active users. It’s not trading volume. It’s not revenue,” because optimizing near-term profit could destroy the user growth carrying long-term value.
CZ portrays Binance’s compliance failure as inadequate registration and controls, not knowing or personal facilitation of illicit transactions. He traces the warning to a Homeland Security request around New Year’s Day 2018, later acknowledges that Binance served US users without proper registration and had weak KYC/AML, and says the court rejected two proposed enhancements alleging his personal facilitation. After the government sought 36 months—twice what he called the maximum guideline—he received four months and entered prison on May 30, 2024.
Binance’s brief FTX investment was strategic but adversarial, and CZ categorically denies that its July 2021 exit was connected to FTX’s later problems. Binance took roughly 20% through an equity and BNB/FTT transaction, but CZ says FTX poached VIP staff at five times salary, solicited Binance customers, and bad-mouthed the company in Washington. Binance exited without using its veto over future rounds; CZ says he was a passive investor who never requested financial statements.
CZ’s forward thesis is that AI agents may drive a major future use of crypto, while privacy remains the industry’s missing prerequisite. He expects individuals to command hundreds, thousands, or millions of agents capable of transacting “a million times more than us,” beyond banks’ ability to KYC or service them. Yet he refuses to name a winning payment token and agrees that traceable wallets expose everything from purchases to hotel locations: “Privacy plays a very fundamental role in our society.”
Post-Binance, CZ is allocating capital and time toward free education while deliberately refusing the obvious tokenization play. Giggle Academy targets an estimated 700–800 million illiterate adults and 500 million children outside school, delivered through a free gamified app; despite a community memecoin project donating about $12 million, CZ says issuing a token would blur real learning with farming. His broader prescription is deliberately unheroic: sustain 110–130% effort for 30 years, get lucky, and remember that “money is only one thread” alongside health, family, time, and mental stability.

Deep dive

1. Scarcity in Canada trained CZ to optimize for function

CZ’s father went to Canada for an academic exchange in 1984, first at the University of Toronto and later at UBC. Obtaining Chinese passports took the family several years; visas arrived shortly after 1989, when CZ was 12. He says Tiananmen-era campus debates may have affected him subconsciously, though “consciously, no.”
His father received roughly C$1,000 a month and faculty housing. On the family’s third day in Canada, CZ’s mother—a former math and history teacher with limited English—started minimum-wage work in a sewing factory and remained there for seven to ten years.
CZ began at McDonald’s around age 14, earning C$4.50 despite a stated C$6 minimum wage because the chain had a youth exemption. Vancouver nevertheless became foundational: “My teenage years in Canada were great,” shaping what he calls his generally happy disposition.
McGill began with biology after a respected adult suggested medicine, but one semester settled the question: “No, no more biology.” He switched to computer science, worked throughout school, took C$6,000 from his father and C$3,000 from his sister, then became self-supporting without student debt.

2. Order-execution code became the hidden precursor to Binance

CZ left McGill before graduating when a Tokyo internship at Fusion Systems Japan kept extending. He later obtained an online computer-science degree from the American College of Computer Science because a bachelor’s degree was required to apply for work visas in Japan.
Fusion built order-execution systems for Tokyo brokers, establishing the through-line to Binance. The software made no trading decisions; the work was moving orders with minimal latency. CZ’s self-assessment is restrained: “I wasn’t a programming wizard,” but efficiency-heavy systems suited him.
His optimization stack began with eliminating database lookups, keeping state in memory, reducing calculations, and simplifying pre-trade risk checks. Moving work onto an FPGA could cut the round trip he remembered from roughly 100 microseconds to 20; custom ASICs were less attractive because algorithms changed too often, while even FPGA reprogramming took about 10 times longer than software.
Fusion sold to a NASDAQ-listed company for about $52 million before 2000, but CZ received none of it. Its founders’ next venture, Building 2, rented an extravagant office and generated zero revenue, folding within a year—his early evidence that mergers can fail culturally and that “previous success doesn’t guarantee future success.”

3. Bloomberg taught CZ to scale teams before he attempted ownership

After September 11, CZ called to confirm whether Bloomberg’s New York offer still stood; Bloomberg asked whether he still wanted to come. He arrived in November 2001, when the city remained subdued, and joined the newly assembled Tradebook Futures team.
Bloomberg looked enormous beside his roughly 20-person Tokyo employer: about 3,000 employees, elaborate offices, fish tanks, and free food. He joined as a senior developer, was promoted three times in two years, and grew from leading roughly 60 people to about 80.
That promotion ended his coding career and began management—“the worst transition,” as the hosts joked. He stayed four years, not because of an entrepreneurial master plan, but because he was a 24- or 25-year-old seeking experience, stability, and exposure to another part of the world.

4. His first startup survived by abandoning its original thesis

In 2005, CZ and five colleagues moved to Shanghai to import Wall Street trading technology. The lead partner held roughly 39–40%; CZ and four others split the remainder, leaving him around 11%, though he understood virtually nothing about preferred versus common shares or shareholder protections.
The key discovery came after incorporation: as a wholly foreign-owned enterprise, the company could not serve the Chinese brokers and financial institutions it had targeted. It pivoted into general IT work—anything from fixing printers to SAP implementations—rather than defend an impossible original pitch.
The improvised consultancy eventually served Shanghai General Motors, Shanghai Volkswagen, Shanghai First Automotive Works, and later Morgan Stanley, Deutsche Bank, and Credit Suisse through Hong Kong. It reached roughly 200 employees and remained operating when CZ described it.
CZ reinvested most of his savings and never cashed out, but the business eventually paid partners six-figure salaries sufficient for international-school tuition. He left in 2013 after eight years; the lesson was less venture romance than earning the right to continue.

5. Bitcoin conviction turned into a concentrated balance-sheet decision

At a Shanghai poker game in July 2013, venture investor Ron Tao—then a managing director at Lightspeed China—told CZ to study Bitcoin. Bobby Lee then proposed a clean asymmetric bet: put 10% of net worth into BTC; a zero wipes out 10%, while a 10x doubles total wealth.
CZ spent roughly six months reading the nine-page white paper and BitcoinTalk discussions. BTC rose from about $70 to $1,000 during that study, producing the universal convert’s reaction: “I’m too late.” The counter-framing was equally durable: no matter when one enters Bitcoin, everyone already inside seems earlier.
A December 2013 Las Vegas conference sharpened his conviction. Media coverage after the Silk Road arrest depicted Bitcoin as drug-lord infrastructure; the approximately 200 attendees—including Vitalik Buterin, Matt Rosak, and Charlie Lee—looked to CZ like “a bunch of kids and geeks,” and notably decent people.
CZ decided Bitcoin was the second foundational technology of his life after the internet: “I wasn’t going to miss it.” He sold his Shanghai apartment for just under $1 million, roughly $900,000, and bought BTC with each payment tranche—first around $800, then through $600 and $400, averaging approximately $600.

6. Chamath’s costly Bitcoin purchases exposed the fallacy of mental accounting

Chamath recalled using BitPay to prove Bitcoin’s transactional utility by buying a Range Rover and Lake Tahoe land. At later BTC prices, he jokingly valued the car at $90 million and the property at $1 billion: “I calculate it this way every day.”
CZ rejected the regret calculation as “the bucket of money mentality.” Had Chamath paid with cash, that cash could instead have bought Bitcoin; choosing a consumption asset over BTC was the economic decision regardless of which payment rail settled it.

7. Two crypto employers supplied the practices—and values—CZ wanted

CZ became roughly the third person at Blockchain.info, working remotely as vice president of engineering while everyone was paid in Bitcoin. The platform had about two million users, and its marketing engine was essentially one 150-page BitcoinTalk thread maintained through founder Ben Reefs’s replies: proof that “guerrilla marketing” could build enormous reach.
Leadership maneuvering changed the culture after the team expanded to around 18 people. CZ left after six or seven months, followed by several developers he had hired and later founder Ben Reef; the remote, crypto-paid operating model nevertheless became a major Binance inheritance.
At OKCoin, competing offers raised his proposed equity from 5% to 10%, and he joined as CTO in Beijing. He departed after roughly eight months over practices he considered misleading, including advertising fee discounts that customers received only if they explicitly requested them.

8. A convincing demo created the recurring-revenue bridge to Binance

In 2015, CZ and two developers planned a Japanese Bitcoin exchange. They downloaded and adapted open-source exchange software, improved the interface, and piped in Bitfinex market data so the order book looked alive; CZ openly called it a demo, but could answer investors’ architecture questions with uncommon depth.
Investors doubted that a non-Japanese-speaking team could operate the exchange, but suggested selling the technology. Within two weeks, a customer signed a $360,000 contract and paid $180,000 upfront—enough to stop CZ personally funding the developers’ salaries.
The team became an exchange-as-a-service vendor, signing about 30 clients over two years. Fixed monthly fees made each new deployment a durable revenue step, producing what CZ described as a multimillion-dollar, highly attractive SaaS business.

9. Customer extinction forced the Binance pivot

In March 2017, Chinese authorities shut most clients, which CZ clarified were primarily stock exchanges and other non-crypto clients, not crypto exchanges. The software company remained untouched, but by May it had almost no customers.
Three employees briefly proposed a Poloniex copycat before, three days later, switching to on-chain arbitrage-trading software. CZ withdrew the investment offer and asked the obvious question: why not operate the exchange themselves, given that the technology and a 20-person team already existed?
The decision was crypto-to-crypto only. The company had technical depth and two salespeople but virtually no consumer marketing—a material reversal from selling infrastructure to businesses into operating a global financial marketplace.

10. BNB financed the exchange and recruited its first users

Li Xiaolai’s June 2017 ICO raised $15 million in ten days from a white paper and website, with no product. After a conference where everyone kept telling him, “CZ, you’ve got to do an ICO,” CZ called his team on June 14 and ordered a white paper.
CZ had enough reputation from Blockchain.info, OKCoin, conferences, and social media to underwrite a new brand. He estimates about 20,000 people bought the ICO, with 80–90% of purchases coming from China, though he says he still does not know the actual buyers.
Binance sold roughly 60% of BNB for $15 million without selling equity. The near-term utility was concrete: holders could receive a 50% trading-fee discount; the white paper also envisioned a native chain and decentralized ecosystem.
On September 4, CZ recalls seven Chinese departments prohibiting crypto exchanges, ICOs, and mining. Roughly 30% of Binance users were Chinese but 70% were elsewhere, so the company cut off the former and moved its approximately 30-person team to Tokyo: “Everybody moved.”

11. Speed and token utility turned launch volatility into hypergrowth

BNB priced around $0.10, fell 30–40% to roughly $0.06, and needed about three weeks to recover. That decline created immediate pressure despite an exchange product CZ says already had market fit.
BNB gave ICO participants a reason to trade on Binance, but execution quality reinforced the incentive: “Using your eyeballs,” customers could see that order placement was faster than on competing platforms such as Poloniex and Bittrex.
Revenue soon reached a few hundred BTC, prompting repeated checks because CZ assumed the number must be wrong by an order of magnitude. During BNB’s recovery, he recalls sleeping, attending a meeting, or visiting the bathroom and returning to find the token another 20% higher.
Forbes featured him in early 2018, roughly six or seven months after launch. He accepted the photo shoot—and his first makeup—not as personal validation but because a still-new Binance brand could use the awareness.

12. Daily active users—not volume—became Binance’s governing metric

CZ rejects the idea that he became addicted to growth: “I was addicted to the work.” A typical day held more than 20 scheduled meetings or calls, additional operational problems, and Twitter responses; the reward was fulfillment rather than the money itself.
His north star was “daily active users. It’s not trading volume. It’s not revenue.” A product has value when people voluntarily use it, even at zero revenue; optimizing short-term profit can weaken the growth that creates value for both the platform and its customers.
Wealth arrived too abruptly to create a consumption ladder. CZ went from barely financially free to a Forbes cover without progressing through cars, yachts, and other upgrades; during the China-to-Japan move, even upgrading a red-eye flight from economy to business required a practical sleep justification.
His old house still developed living-room leaks, but it fit the family and location. The explanation was not asceticism: “I’m function-driven.” His emotional “amplitude” is similarly narrow—he experiences happiness and sadness without the extreme swings he sees in others.

13. Compliance became unavoidable on New Year’s Day 2018

Around New Year’s Day 2018, a US Homeland Security officer requested help tracing funds from the EtherDelta hack. Binance first verified his identity, supplied the requested information, and asked him to recommend someone who could handle future law-enforcement contacts.
The recommended candidate was US-based, but Binance then lacked an American entity and could not hire him. For CZ, that message made the scaling problem explicit: some daily active users would be bad actors, and the company needed specialists who understood law enforcement.
By 2019, reports involving BitMEX and frozen assets associated with Bitfinex convinced Binance it needed registered US operations. Binance.US was established as a separate entity with its own deployment, matching engine, and liquidity; CZ says it was regulated from day one.

14. Binance’s FTX investment quickly became an adversarial partnership

CZ first met Sam Bankman-Fried in January 2019, when Alameda was a large Binance trader. Binance rejected proposed futures-platform partnerships, including one CZ remembers as 60/40 in Binance’s favor, because Binance already had the users and saw little need for the joint venture.
After FTX launched, Binance accepted an attractive deal for roughly 20% equity, involving a BNB/FTT token swap. BNB was substantially more liquid; FTT was not, making Binance both shareholder and competitor.
Friction appeared almost immediately. CZ says FTX paid five times Binance salaries to hire a VIP account manager, whose move was followed the next day by calls offering Binance’s VIP clients better rates; he also repeatedly heard that SBF was bad-mouthing Binance in Washington.
Binance exited in July 2021 without exercising contractual veto rights over future rounds. CZ says he never requested financial statements because he was passive and wanted FTX to operate independently; claims that his exit was connected to FTX’s later problems are “categorically not true,” while an ongoing lawsuit limits his bankruptcy commentary.

15. Enforcement negotiations became a year-long test of asymmetry

US information requests began around 2021–2022, according to CZ, and Binance complied. The posture grew more hostile in late 2022; by early 2023, the choice appeared to be reaching a settlement or facing indictment. CZ also characterizes the Biden administration as hostile to crypto, while noting that legal restrictions limit what he can say about the plea.
The hardest operational problem was coordinating 12–20 expensive lawyers with different specialties and opinions through daily calls lasting more than a year. “We have never seen them this hostile on a case like this” became, he says, the phrase he heard most often.
CZ mapped outcomes from a fine or deferred-prosecution agreement to jail, a Red Notice, and permanent confinement within the UAE. Fighting from a non-extradition country might have protected him from extradition, but restricted travel and burdened a government that had granted him citizenship: “I don’t want to be the trouble causer.”
Several times Binance rejected proposed terms and heard nothing for about two weeks. CZ feared a sealed indictment at every border; when talks resumed, he came to see the silence as calibrated pressure—long enough to create fear, but not long enough for the target to normalize the new life and refuse further concessions.

16. The plea separated control failures from intentional facilitation

With the explicit caveat that he is not a lawyer, CZ describes his single charge as a Bank Secrecy Act registration failure: Binance served US customers without registering properly. He also says the company’s KYC/AML program was weak.
His hierarchy matters: inadequate controls can miss bad actors without anyone knowingly helping them; a more serious level is knowingly facilitating illicit transfers; beyond that is personally handling them. “I don’t handle any transactions at all myself,” he says.
The government sought two enhancements alleging personal facilitation, but CZ says it could not identify supporting transactions and the court rejected both. He entered the process expecting that no one had gone to jail for this type of registration violation, citing Arthur Hayes’s six months of home confinement as the harshest precedent he knew.
After his plea, a magistrate initially allowed him to await sentencing in the UAE, but the government appealed and kept him in the US. Before the April 30, 2024 hearing it requested 36 months—twice what CZ called the maximum guideline—while his team sought probation or home confinement; the judge imposed four months.

17. Prison’s uncertainty mattered more than the four-month term

The sentence itself was less frightening than physical uncertainty. Media called CZ the richest person ever headed to US prison, while consultants described him as a prime extortion target and warned against accepting friendly favors that might later carry violent repayment demands.
CZ entered on May 30, 2024. Although his offense would normally qualify for minimum security, he says his non-US citizenship put him in a low-security prison holding about 2,200 inmates; his unit contained roughly 200 men.
A guard immediately suggested protection through the Pacific Islanders. Inside, inmates organized into racial or ethnic “cars,” with representatives resolving intergroup disputes; an Asian-looking inmate introduced himself as Chino and said, “Welcome to our car,” leaving CZ unsure whether shaking hands meant joining a gang.
No physical harm, fight, or meaningful extortion ultimately occurred. The lasting luxuries were basic: a shower without touching the walls and whole fruit after months of carb-heavy meals with little produce or protein. On September 27, he says, about six hours and 26 minutes separated leaving custody and boarding a flight.

18. Losing Binance created room for education, investing, and a pardon

Stepping down was emotionally difficult: CZ cried, something he otherwise remembers doing only when his father died. Later, the prohibition became liberating—he could not be accused of losing stamina, and he discovered “other very useful, meaningful things” to do.
He describes the pardon process as opaque: a lawyer writes a petition explaining why the applicant deserves relief, sends it into a system with a pardon czar, repeatedly requests updates, and waits. CZ says he “didn’t do much” beyond that process.
His explanation for why a pardon made strategic sense is explicitly conjecture. As ultimate beneficial owner, his conviction constrained Binance and Binance.US; he reasons that a president seeking US crypto leadership would want citizens to access the largest player and one of the largest crypto ecosystems, and might sympathize after experiencing aggressive prosecution personally.
Outside Binance, CZ says YZi Labs invests across blockchain, AI, and biotech. He also consults governments on crypto regulation and mentors founders in the BNB Chain ecosystem.

19. Giggle Academy rejects token incentives to protect learning integrity

CZ frames the addressable education deficit as 700–800 million illiterate adults—roughly two-thirds women—plus about 500 million children outside school, or around 1.2 billion people. His proposed delivery system is a free app on phones or tablets.
AI, gamification, psychology, and digital content could, in his view, deliver all the education content people need without building schools. He admires Alpha School but distinguishes its expensive mission of improving existing education from Giggle’s initial goal of making any education accessible.
A token would enable “learn to earn,” reward teachers, and stimulate content creation, but CZ intends to resist one “for a very long time.” Given his profile, speculation would make it impossible to distinguish children learning from farmers maximizing token rewards: “I want Giggle to be a real free education platform.”
Funding is not yet the constraint. A community memecoin project donated roughly $12 million, while CZ estimates total spending at only $3–4 million; he plans to finance the platform as long as needed, while acknowledging that giving money away with positive impact is “very, very difficult.”

20. AI agents may need crypto before humans make it ubiquitous

CZ calls AI the third foundational technology of his life after the internet and Bitcoin. He expects each person eventually to command hundreds, thousands, or millions of agents working in the background, buying services and moving money.
The volume argument is central to the discussion: “The agents can transact a million times more than us.” Banks may not sensibly KYC an agent or support the speed and granularity of machine commerce, while crypto is discussed as a possible payment rail for programmatic purchases such as tickets, restaurants, hotels, or podcast access.
Trading interfaces should also disappear. Instead of reading charts and entering orders, a user might say, “Convert 10% of my stablecoins into BNB”; the agent would choose one market order for a small position or execute gradually for a large one.
Asked which existing crypto project is best positioned, CZ gives an honest non-answer: it is early, many teams are working on it, and naming one could move its token price. The thesis is forward-looking; the investable winner is not.

21. Privacy remains crypto’s unresolved path to everyday money

Jason’s objection to Bitcoin maximalism is twofold: insufficient fungibility and insufficient privacy. Dollars reveal nothing about prior owners or purchases, while public-chain histories can make one unit economically or reputationally distinct from another.
CZ agrees that Bitcoin’s intended pseudonymity is inadequate in practice. Every transaction is traceable, and centralized-exchange KYC makes wallet identities easier to connect, leaving “Bitcoin and most cryptocurrencies” without enough privacy.
The strongest example is physical security: paying a known hotel address can reveal where someone will be staying. Privacy is not merely protection for illicit activity; ordinary purchases, entertainment, cigarettes, travel, and home addresses need not become permanent public disclosures.
CZ supports law enforcement’s ability to pursue bad actors but argues that the industry must evolve privacy features simultaneously. He corrects himself that some privacy-focused coins exist, yet notes their low market capitalizations and limited scale.

22. CZ’s closing doctrine is endurance without billionaire mythology

CZ began his roughly 95,000-word book on a “very dumb terminal” in prison, sending drafts to his assistant. Editing English and Chinese versions takes weeks per pass, but the project now serves to explain misconceptions about CZ, Binance, and crypto—and to give his children detail he never had time to tell them.
His parental ambition is deliberately open-ended: healthy, happy children, whether they choose ordinary work, startups, art, or humanitarian projects. It echoes what his own parents told him: “Don’t hurt yourself. Don’t hurt other people.”
CZ’s success prescription rejects genius worship: “I’m a normal dude.” Principles, values, emotional intelligence, and luck matter alongside ability; push sustainably at roughly 110–130%, not toward burnout, and compound that effort for 30 years. The likely result may not be billions, but could be a comfortable life.
Wealth is only one thread in a wider web. Once basic needs are secure, more money contributes less than health, family, time autonomy, meaningful work, and mental stability; chasing it while sacrificing everything else defeats the point. That is why CZ can now say, “I’m actually very grateful that I don’t have to run Binance anymore.”

The Secretive PE Firm Behind Burger King, Tim Hortons, Skechers and Hunter Douglas (3G Capital)Market Overview, February 10, 2026