Transcript
0:00 Mark Zuckerberg renamed his company, bet its future on a virtual world, and lost more than $80 billion doing it. In 2022, Meta's stock crashed more than 60% and Wall Street called him reckless. Today, he is worth more than $200 billion. So what happened in between? Here is the strange part. The metaverse never took off the way he promised. Almost nobody lives in it, and yet Zuckerberg came out of this richer and more powerful than ever. The story of how he did it says a lot about where AI and your money are heading next. We will go through it in order. The big announcement, the crash, the layoffs, the pivot nobody saw coming, and the new bet that is even bigger than the first one. All numbers come from Meta's own financial reports and public reporting listed in the description. But first, to understand why he would make a bet this big, you have to remember who Mark Zuckerberg is.
1:10 In February 2004, as a 19-year-old student at Harvard, he launched a website called The Facebook from his dorm room. Within a month, more than half of Harvard's undergraduates had signed up. In 2006, Yahoo offered to buy Facebook for about $1 billion. Zuckerberg was 22. His investors and many of his employees wanted him to take it. He said now that decision is the key to everything that came later. He has always believed he can see further than the people around him. He was often right. In 2012, he bought Instagram for about $1 billion when it had just 13 employees. People called crazy. Today, Instagram has billions of users. In 2014, he bought WhatsApp for about $19 billion.
2:10 Again, people said he overpaid. Facebook went public in May 2012 at $38 a share, raising about $16 billion. The first months were rough, and the stock fell by about half. But Zuckerberg moved the company to mobile, and the stock recovered. Every time he made a big, unpopular bet, it seemed to work. And he kept total control. Thanks to special voting shares, Zuckerberg controls a majority of Meta's voting power, even though he owns a much smaller share of the stock. Nobody can fire him. So when he decided to bet the company on the metaverse, there was no one who could stop him. October 28th, 2021.
3:02 Facebook's annual Connect conference. Mark Zuckerberg walks through a virtual living room and announces that the company called Facebook will now be called Meta. He says the metaverse will be the successor to the mobile internet. He did not stop at a new name. He said that from now on, the company would be metaverse first, not Facebook first. The idea was a 3D internet where you could work, play, and hang out as an avatar using virtual reality, he had said. And one day, glasses. Why would he do this?
3:41 Facebook was under huge pressure at the time. A whistleblower, Frances Haugen, had just leaked internal documents to the Wall Street Journal and testified to Congress. Apple had changed iPhone privacy rules in a way that hurt Facebook's ad business. Zuckerberg wanted a platform he controlled himself. It was not the first crisis. In 2018, the Cambridge Analytica scandal revealed that data from up to 87 million Facebook users had been collected by a political consulting firm. Zuckerberg testified before Congress for two days. In 2019, the Federal Trade Commission fined Facebook $5 billion over privacy. And young people were leaving. TikTok was exploding and teenagers saw Facebook as an app for their parents. Internal research leaked in 2021 suggested Instagram made body image issues worse for some teen girls.
4:50 The brand Facebook had become toxic in Washington and boring to teenagers. That last point matters. For years, Zuckerberg had complained that Facebook lived on phones made by Apple and Google, so those companies set the rules. If the next computer was a headset made by Meta, nobody could do that to him again. The metaverse was his escape plan, and he was ready to spend. Meta had already bought Oculus in 2014 for about $2 billion. Now its virtual reality division called Reality Labs was given a budget measured in tens of billions of dollars a year. But here is what happened to that money. In 2021, Reality Labs lost about $10 billion.
5:44 In 2022, about $13.7 billion. In 2023, about $16 billion. In 2024, about $17.7 billion. And in 2025, Reality Labs lost about $19.2 billion dollars a single year. Add it all up since late 2020 and the losses pass 80 billion dollars. That is more than the entire value of many famous companies.
6:16 did that money buy? like the Quest 2 and Quest 3 sold millions of units and became the best selling VR devices in the world, but VR stayed a niche, mostly for gaming and fitness. Most people tried it, put it in a drawer, and went back to their phones. Meta's own social world, Horizon Worlds, struggled to keep users. Internal documents reported by The Wall Street Journal in 2022 said most visitors did not come back after the first month and that the company was well short of its user targets. Then came the moment the internet never forgot. In August 2022, Zuckerberg posted a selfie of his avatar in front of the Eiffel Tower and a Spanish cathedral to launch Horizon Worlds in France and Spain. The cartoon graphics were mocked around the world. A few weeks later, Meta announced that avatars would finally get legs. Then it had to admit the legs shown in the demo were made with motion capture, not with the real product. For many people, the metaverse had become a joke. To put 80 billion dollars in perspective, it is more than NASA budget for three years.
7:41 It is more than Meta paid for Instagram and WhatsApp combined, four times over, all spent on a world that most people have never visited. And Wall Street stopped laughing. On February 3rd, 2022, Meta reported that Facebook had lost daily users for the first time in its history. The stock fell about 26% in one day. Meta lost more than $230 billion of value, than the biggest one-day loss for any American company.
8:17 It kept falling. By early November 2022, Meta's stock was below $90 down from almost $380 a year earlier. The company that once was worth more than a trillion dollars was suddenly worth around $250 billion. Investors were furious. In October 2022, Brad Gerstner, the head of Altimeter Capital, a Meta shareholder, published an open letter to Zuckerberg. He wrote that Meta needed to get fit and focused, and he asked Zuckerberg to cut metaverse spending to no more than $5 billion a year. On Meta's earnings call of October, Zuckerberg did not back down.
9:06 He acknowledged that many people disagreed with the investment, but said he believed it would be very important and that the company would be glad it made the bet. The stock fell another 20% after that call. So here is the question at this point in the story. Zuckerberg controls the majority of Meta's voting power, so no one could force him to stop. If the metaverse was failing, how was he going to save his company? The first answer was painful. In November 2022, Meta laid off about 11,000 employees, roughly 13% of its workforce. It was the first mass layoff in the company's history. Zuckerberg wrote that he got the growth wrong and took responsibility for it. Then in February 2023, he told investors that 2023 would be the year of efficiency. In March, Meta announced about 10,000 more job cuts and closed thousands of open positions. Layers of managers were removed. Wall Street loved it. Costs fell, profits rose, and the stock started to climb back. But efficiency alone was not a strategy.
10:28 Zuckerberg needed a new story, and it came from somewhere almost nobody expected. Here is the twist. While the world was mocking the metaverse, Meta had been building one of the best AI research labs in the world for years. It was called Fair, and it was led by Yann Le Cun, one of the godfathers of deep learning, who won the Turing Award with Jeffrey Hinton and Joshua Bengio. Meta had also bought huge numbers of Nvidia chips. Zuckerberg later explained that the company had ordered them to power the recommendation engine for Instagram Reels to compete with TikTok. When ChatGPT arrived, Meta already had the computing power to compete. In February 2023, Meta released LLaMA, its own large language model. In July 2023, it released LLaMA 2 for free for research and commercial use with Microsoft as a partner. While OpenAI and Google kept their models closed, Zuckerberg gave his away. Why give away something so expensive? Zuckerberg wrote that open source AI would become the industry standard, the way Linux did for servers, and that it would stop any single company from controlling the technology.
11:57 It was the same lesson he learned from Apple applied to AI. Market reaction was enormous. Meta stock rose about 190% in 2023, one of the best years for big company. The story had changed. Meta was no longer the Metaverse company. It was an AI company. Quick pause. If you want to understand real stories behind Zuckerberg, Musk, Altman, and the billionaires shaping AI, subscribe to 2100.
12:33 There's a new one every day. Now, because the comeback did not stop there, something else surprised everyone. The part of the Metaverse plan that actually worked was not the headset. It was glasses. Meta's Ray-Ban smart glasses made with EssilorLuxottica have a camera, speakers, and an AI assistant built in, and they look like normal sunglasses. EssilorLuxottica said that Meta glasses sales tripled in the first half of 2025.
13:09 In September 2025, Meta launched the Ray-Ban display glasses with a small screen inside the lens controlled by a wristband that reads tiny muscle movements. Not everything went smoothly. At the 2025 Connect launch, several live demos of the new glasses failed on stage. A video call would not connect, and a cooking assistant gave the wrong steps. Zuckerberg laughed it off and blamed the Wi-Fi. The clips went viral, a reminder that even trillion-dollar companies still struggle with live demos.
13:49 Zuckerberg now says glasses will be the main way people use AI. He has said that in the future, people who do not have AI glasses could be at a significant cognitive disadvantage. That is a bold prediction, and history is full of smart glasses that failed, including Google Glass. So the Metaverse money was not completely wasted. Much of the technology from displays to sensors to the wristband went into the glasses. And Reality Labs is moving most of its spending toward wearables. The question is whether glasses can ever earn back a billion dollars. There was a rivalry inside this story too. In 2023, Apple announced its own headset, the Vision Pro, priced at $3,499.
14:45 Tim Cook carefully avoided the word metaverse. Apple called it spatial computing. In February 2024, Zuckerberg posted a video reviewing the Vision Pro himself. He said the Quest was not just the better value, but the better product, period. It was rare for a CEO to review a rival product so directly. He clearly saw this as a fight against Apple for the next computer. The Vision Pro did not become a mass market hit either, and Apple reportedly scaled back production. Both giants learned the same lesson. People were not ready to wear a computer on their face for hours.
15:30 The winner, it seemed, would be whoever made something light enough to wear all day. In September 2024, Meta showed a prototype called Orion, true augmented reality glasses that project images onto the world in front of you. Meta said each pair costs so much to build that it was not ready for sale, but early testers said it felt like the future. But Zuckerberg was not done. In April 2025, Meta released Llama 4, and many developers said it disappointed. Meta was accused of tuning a version of it to look better on a popular benchmark.
16:15 Rivals like OpenAI, Google, and the Chinese lab DeepSeek seemed to pull ahead. Zuckerberg's answer was to spend. In June 2025, Meta invested about $14.3 billion in Scale AI for a 49% stake and hired its founder, Alexander Wang, to lead a new group called Meta Super Intelligence Labs. Then he went shopping for researchers. Sam Altman said on a podcast that Meta had offered OpenAI employees signing bonuses of $100 million.
16:56 Zuckerberg personally contacted top AI scientists and Meta hired dozens of them from OpenAI, Google, and Apple. In July 2025, Zuckerberg published a letter about what he called personal super intelligence. He wrote that he wanted to give everyone a super intelligent assistant that helps them achieve their goals, and he contrasted it with rivals who, in his view, wanted AI to automate work and replace people. And to now look at the size of the new bet, Meta plans to spend around $70 billion on capital expenditures in 2025, mostly AI data centers, and said spending would grow even more in 2026.
17:47 That is close to the entire metaverse loss every single year. Meta is building a data center in Louisiana called Hyperion, which Zuckerberg said could eventually scale to 5 gigawatts with a footprint that would cover a large part of Manhattan. Another called Prometheus is planned in Ohio. These are some of the biggest buildings ever planned for AI. Zuckerberg has said he would rather risk misspending a couple of hundred billion dollars then be late to super intelligence. That sentence tells you everything about how he thinks. He learned from the metaverse that he can survive a huge mistake as long as he owns the company. And in late 2025, Bloomer reported that Meta planned to cut its metaverse budget by up to 30%.
18:43 The money was moving to AI and glasses. Quietly, the metaverse first era was ending four years after it began. Washington also tried to break Meta up. The Federal Trade Commission sued, arguing Meta bought Instagram and WhatsApp to crush competition, and the case went to trial in 2025. In November 2025, a federal judge ruled in Meta's favor, finding the government had not proved Meta held an illegal monopoly today, in part because of competition from TikTok and YouTube. That ruling removed one of the biggest threats to the company.
19:29 Zuckerberg could keep Instagram and WhatsApp, the apps that generate much of the money funding his AI bet. Again, the core business survived, and the core business is what pays for everything. Zuckerberg also changed his public image. He started competing in Jiu-Jitsu tournaments, grew his hair out, wore gold chains and oversized t-shirts, and appeared on long podcasts with Joe Rogan. In January 2025, he attended President Trump's inauguration alongside Bezos, Musk, and Pichai, which brings us to today. This month, Fortune reported that Zuckerberg's fortune passed $200 billion.
20:16 Meta's advertising business, powered by AI that decides which ads you see is making record profits. The stock is far above where it was before the metaverse crash. So how did he win after losing $80 billion? The honest answer is that the core business paid for everything. Facebook, Instagram, and WhatsApp reach billions of people every day. The ads on those apps generated so much cash that Meta could afford to lose tens of billions a year and still be hugely profitable. And AI made that core business better. Meta has said AI recommendations increase the time people spend on Facebook and Instagram, and AI tools helped advertisers get better results. In a way, the AI pivot did not just save the company. It made the old business stronger. Is there a pattern here?
21:17 History said yes. Jeff Bezos lost money on the Fire phone, and Amazon kept going because Amazon Web Services and retail were strong. Google spent years on projects that failed. Company with a cash machine at their core can afford spectacular mistakes. The second pattern is control. Zuckerberg holds most of Meta's voting power. Elon Musk at Tesla and SpaceX and Larry Page and Sergey Brin at Google built similar structures.
21:52 Founders who cannot be fired can make bets no normal CEO would survive. The third pattern is timing. The metaverse came too early. The headsets were heavy, the graphics were simple, and there was no killer reason to use it. AI arrived at exactly the right time. The same leader with the same appetite for risk got one bet wrong and one bet right. But here is the warning. If the AI spending boom does not pay off, the losses could be far bigger than the metaverse. $70 billion a year on data centers is a much larger bet than $20 billion a year on headsets.
22:37 Burry and other skeptics say the AI boom looks like past bubbles. There is one more pattern from history. When a bet fails, the best companies take the parts that work and use them somewhere else. Amazon's Fire phone failed, but its team helped build the Echo and Alexa. Meta's metaverse failed, but its hardware and AI teams built the glasses and the chips that now power its AI. And critics have not gone away. Some former Meta employees and investors argue the company wasted years and tens of billions of dollars that could have gone into AI earlier when OpenAI was pulling ahead. Others argue the reverse, that the metaverse spending built a culture willing to make huge bets. What does this mean for you? First, the AI in your fee is getting smarter and it is learning what keeps you scrolling and what makes you buy. Second, AI glasses are coming whether or not you want them.
23:45 And third, a few founders now decide where hundreds of billions of dollars go. If you own an S&P 500 fund, you own Meta, Nvidia, Microsoft, and Google, the companies spending the most on AI. Their bets are your bets too. This is not financial advice, but it is worth knowing. And there is a privacy question. AI glasses with cameras on millions of faces mean much more of daily life could be recorded. Meta says its glasses have a light that turns on when recording. Privacy groups say that is not enough. It is a debate that will only get louder as the glasses get better. Finally, look at jobs. Meta cut about 21,000 jobs during the year of efficiency and then hired some of the most expensive AI researchers in history. That is a preview of the AI economy. Fewer people overall and enormous pay for the few with the rarest skills. So last recap. Zuckerberg renamed Facebook, bet everything on the metaverse, lost more than $80 billion, saw his stock crash more than 60%, cut about 21,000 jobs and then pivoted to AI, classes, and super intelligence.
25:13 Today he is worth more than $200 billion. Was the metaverse a disaster or the expensive training run that made the AI comeback possible? And will the super intelligence bet end like the metaverse or like the AI pivot? Tell me what you think in the comments and tell me if you have ever actually used the metaverse.