This week, Mark Zuckerberg published a 6,500 word manifesto about how Meta is going to deliver personal superintelligence to every human being on Earth. In it, he describes AI tutors with PhDs in every subject, superintelligent lawyers, personalized baking assistants, and a future so abundant that everyone will have invention superpowers. That is a truly inspiring vision coming from a man who spent $80 trying to turn humanity into legless avatars, only to end up creating a glorified VR waiting room for crypto bros and corporate HR departments. I mean, there is something to be said about this man's vision, no? The same week, his $300 million yacht reportedly ignored a distress call from a stranded boat in Alaska. His CTO told an employee to go ask their parents what they think of asking for time off, and his company reported record revenue while free cash flow dropped 91%. That means everything they earned, they spent on new screws for the new data centers. This is a company that has tried to be a metaverse company, a frontier AI lab, a smart glasses company, a geopolitical weapon against China, and based on this manifesto, a personalized baking assistant. There is, however, one thing Meta has not tried to be with the same enthusiasm, the best possible version of the advertising company it already is. We are watching this company decline in slow motion. Everyone can see it except the actual people running it. Don't be fooled by the stock chart. A line can look high, but if you integrate under the curve, all you find underneath is $30 million of burned cash and talent drain the size of a sinkhole swallowing Silicon Valley. The manifesto is called The Future Is for Everyone. It promises that everyone will have a personalized tutor and coach with a PhD in every subject. Having a PhD in a subject makes you extremely specialized in one specific area. It doesn't necessarily make you the king of that entire field. I have a PhD. I can assure you it mostly makes you tired. But sure, a PhD in every subject. Let's continue. It promises super intelligent lawyers for everyone. From the CEO of a company that just paid $567 million in a child safety case. Apparently, the super intelligent lawyers will be available to the public slightly before they're available to Meta's own legal department. It describes a dynamic auction mechanism for compute pricing, which is surge pricing for intelligence basically. Uber, but for your brain. More people thinking at 5:00 p.m., prices go up. Quiet Tuesday, discount reasoning. It declares that there is no such thing as a singular benevolent superintelligence. The CEO who controls a platform used by 4 billion people on this planet just declared that benevolent omniscience cannot exist. That is an interesting theological position from someone who operates the closest thing to it. He's basically telling us, "Look, relax. Nobody can ever be a benevolent god." Which is the ultimate humble brag from a guy who runs a monopoly with the social awareness of an algorithm trying to sell weight loss tea to teenagers. And of course, my personal favorite, Zuckerberg writes that my eight-year-old daughter can already code her ideas and produce videos in an evening. His eight-year-old plans personalized recipes for us to make together each weekend, orders the ingredients, and then offers suggestions as we're baking. The CEO of a company that just fired 8,000 people and spent $1.18 billion on severance is asking AI to plan baking with his daughter. I'm really glad that somebody at Meta is having a good time. It takes a special kind of billionaire sociopathy to lay off thousands of families via automated email and then immediately write 6,000 words of bragging about how an AI assistant bought your organic flour for your weekend scones. Now, I want to be fair here. Anyone can imagine a future. Anyone can write about what super intelligence might do someday. I also have such fanciful dreams, although mine usually involve a machine that can explain Mark Zuckerberg's fashion evolution without using the word crisis. And some of the vision in this manifesto is generally exciting if you take it at face value. But vision without methodology is just daydreaming. Where are the university collaborations? Where are the fundamental research breakthroughs? Where is the substance behind the rhetoric? If this entire planet became a data center tomorrow, the research is still not there for what Zuckerberg is describing. We do not have sentient AI. Generative AI is still producing significant errors. The reasoning capabilities, the nuance, the empathy, virtue's not there yet. And a 6,500 word manifesto does not accelerate the timeline. Research does. Investment in actual science does. Less manifesto, Mark. More methodology. Zuckerberg doesn't want to run a company. He wants to play populist with real people's lives while wearing a gold chain and pretending he discovered streetwear.

And this is where the slow decline starts to actually show. The manifesto is 6,500 words about super intelligence and zero words about advertising. The thing that made Meta the most dominant advertising platform on this planet, the thing that is about to overtake Google for the very first time, doesn't merit a single paragraph. The manifesto does, however, include a section about community benefits in Richland Parish, Louisiana, where teachers received a $50,000 bonus from data center tax revenue. This is the same Louisiana where Meta's contractor contaminated a city's water supply with industrial bacteria, which is of course a detail the manifesto conveniently does not mention. And for someone who has been compared to an alien by roughly the entire internet, I'm talking about the sunscreen surfing incident, the congressional testimony where he appeared to be experiencing human hydration for the first time, the dead-eyed that has launched a thousand memes, but really let's not trash his looks too much. We're all stuck with our phenotypes. What I'm trying to say is writing a manifesto about empowering humanity feels like a choice, a bold one. The kind of choice you make when you're either deeply sincere or deeply unaware of how other people perceive you, and at this point I genuinely cannot tell which it is. I mean, the guy owns a diamond mine and he's too embarrassed to go down the shaft because it's dirty and you might get cooties down there. So, instead he writes a manifesto about how one day diamonds will fall from the sky. Meanwhile, the actual diamonds are right there in the ground waiting on mined. Still, if there is one thing this manifesto accidentally gets right, it's that education matters. The gap between imagining a super intelligent future and actually understanding how any of this works is enormous, and closing it requires learning, not manifestos. Coursera is sponsoring today's video, and they offer a huge range of courses covering generative AI, prompt engineering, automation, and AI development. I've tried several of these courses myself, and what I particularly like is that they're not all designed for the same kind of learner. You can start with the fundamentals, focus on using AI more effectively in your existing job, or even move into building AI-powered applications yourself. Google AI Essentials is a strong starting point for beginners because it focuses on practical ways to use generative AI in everyday work. It covers things like improving productivity, generating ideas, working with AI tools more confidently, and understanding how to use them responsibly. Vanderbilt's prompt engineering course goes deeper into how prompts actually shape the quality of an AI system's output. It introduces more structured prompting techniques that can help you produce clearer, more useful results, and begin using generative AI for more complex workflows and automation. And for people who want something more technical, IBM's AI developer professional certificate is much more hands-on. It explores how AI applications and chatbots are built, giving you the opportunity to develop practical skills that can be applied to automation and real business problems. There also courses covering AI ethics, large language models, and generative AI engineering, so you can really choose something that matches both your current experience and where you want to go next. I'll leave links to these courses, along with a few more options, in the description below. Those are affiliate links, by the way, which means I may earn a commission if you enroll at no additional cost to you. So, whether you want to understand AI, use it more effectively, or start building with it, definitely explore Coursera's AI courses using the links below. Right, back to someone who could probably benefit from a course in situational awareness.

The Launchpad is becoming a recurrent character in the meta story, and not in a way that any PR team would actually choose. In May, the $300 [million] 387-ft superyacht sailed into Seattle the same day Meta cut a thousand 4,000 local jobs. People booed from the shore. Now in August, the Launchpad was cruising Southeast Alaska when a 21-ft skiff ran out of fuel between Petersburg and Juneau. The Coast Guard broadcasted a call for assistance. Zuckerberg's yacht, which was closer, did not respond. A 42-year-old cruise ship called the Wilderness Legacy, which is a vessel roughly 1/10 the value of the launchpad, steamed past the superyacht and towed the skiff to safety. Passengers on the cruise ship booed Zuckerberg when the captain announced what had happened. I mean, imagine being so universally disliked that tourists on a budget cruise in Alaska will pause their vacation just to boo your empty $300 [million] boat. It sounds like we might have reached global consensus levels here, Mark. I'm really sorry about that. Maritime law expert Ian Ralby told the press, and I'm quoting here, "Human beings are required to save the lives of other human beings if they can do so at sea." I can't even believe that we have to say this kind of thing out loud. I am regularly horrified by how little empathy this planet appears to have. Can we please, for once, do our prefrontal cortices justice? Seriously. Of course, Zuckerberg's spokesperson said the crew didn't hear the call. Zuckerberg was not on the boat. The launchpad is a $300 [million] bad omen that sails around the American coastline leaving PR disasters in its wake. Does this man have any situational awareness at all? I mean, you fire 1,400 people in Seattle, yacht shows up. Someone needs help in Alaska, yacht doesn't respond. I mean, no wonder he wants a bunker somewhere. He is building an apocalypse compound in Kauai because he knows if the grid ever goes down, the first thing humanity is going to do is ask him for a refund on the last 15 years of social media. But the yacht isn't just bad optics. It is a symbol of a much deeper problem. A leadership so detached from reality that it doesn't notice what's happening around it. Not the stranded boats, not the employee morale, not the 91% free cash flow collapse, not the fact that the diamond mine is right there and nobody is digging.

Also, Meta has now tried to be a cloud provider, by the way. On the earnings call, Zuckerberg hinted at building a cloud business to compete with AWS, Google Cloud, and Azure. He, of course, provided no product description, no timeline, and no competitive positioning. The market read it as more spending, less clarity. Let's just add that one to the list of shenanigans. For a second, let's look at their finances and see what's actually going on. I'm going to review some numbers before we discuss their conclusions. Their revenue was $60.8 billion, up 28%. That is a record for them. The advertising engine continues to be one of the most effective money-printing machines in the history of commerce. Capital expenditure, $31.1 billion in a single quarter, nearly double the year before, almost all of it going into AI infrastructure. Free cash flow, $784 million, down from $8.5 billion. That is a 91% decline. The advertising engine generated $31.9 billion in operating cash flow, and the AI build-out consumed 98% of it. Meta spent almost every dollar the business could produce. Earnings per share, $6.18 versus the $7.10 Wall Street expected. Missed, despite record revenue. Legal charges, $2.4 billion. Severance, $1.18 billion. New debt issued in Q2 alone, $25 billion. And while Meta burns cash on AI infrastructure, the AI-generated content flooding its own platforms is driving users away. The AI slop backlash has reached the point where platforms are actively building tools to flag, label, and ban AI-generated content. Medium lost subscribers after telling users they don't really hate AI articles, which is the kind of thing you say when you've stopped listening to the people who actually use your product completely. Meta is simultaneously spending billions to generate AI content and discovering that people don't want AI content. Mark is running a 1.3 trillion dollar corporate feedback loop, spending billions of dollars to force-feed algorithmic slop to users who hate it, while using his own profits to buy the bib. The left hand is building what the right hand is trying to ban, or the AI giveth and the AI taketh away. Take your pick. Naturally, the stock fell roughly 10% after the report. Record revenue, stock fell, genius Mark. The Next Web summarized it perfectly. They said, "Meta is spending like a cloud provider without being one yet, funding a superintelligence ambition out of an advertising engine, and asking shareholders to trust that the two halves eventually meet." I'm not saying Meta will crash tomorrow. The ad engine is still making money, but that engine is funding an expedition to find different diamonds somewhere else, while the actual diamond mine sits untouched. Every quarter where 98% of the cash flow goes into AI infrastructure that produces no independent revenue is another quarter of not mining what's already there, and diamond mines do not wait forever. Competitors learn, data advantages erode, and one day you look up from your superintelligence manifesto and discover that somebody else built the advertising AI you should have built, using the opportunity you left sitting on the table while you were writing about baking with your daughter. I've nothing against his daughter, by the way. I'm sure she's a cutie. Anyway, moving on to the Victorian time. During a July Q&A with staff, an employee asked Meta CTO Andrew Bosworth whether AI productivity gains could be used to revive Meta days, which is a canceled holiday program that once gave staff additional days off. Bosworth's response was, "I hope that what we do with our extra for is do even more and cooler stuff for the users who use our products every day. Of course, we were all born to maximize shareholder value and please our users. What a dumb question." When the employee pressed Bosworth, he insulted them. "Go to your parents and ask them, hey, like every time I get a chance to talk to my boss, ask me if I can have more days off. Ask your parents what they think of that as a career strategy." He later apologized for coming down too hard and called it tongue-in-cheek. Sure. The same week, OpenAI CEO Sam Altman told a podcast he doesn't expect AI to lead to a shorter work week because workers are secretly going to be happy about being much busier. The manifesto says the future is for everyone. The CTO says the future is for everyone to work harder and stop asking for time off. Between Mark's baking fairytales and Bosworth's Victorian factory owner management style, Meta's internal culture reads like a Dickens novel adapted by Silicon Valley venture capitalists. "Please, ma'am, may I have more PTO?" This is the same CTO who admitted morale was near 20-year lows and whose solution was increased snack budgets. This is a leadership team that has fired 8,000 people, spent 1.18 billion dollars on severance, told the survivors to have fun at a hackathon, berated them for wanting a day off, and then published a manifesto about how AI will empower every person on Earth. When leadership gives this kind of direction, the best people leave. And when the best people leave a company whose only real competitive advantage is its data and the humans who know how to use it, the decline accelerates. I'm again not saying that this is going to happen tomorrow, but it is happening quietly, persistently, one resignation at a time while the CEO writes manifestos and the yacht sails to Alaska.

And I spoke with advertising professionals who actually use Meta's tools. The Advantage Plus suite, the targeting systems, the optimization platform. They told me the targeting still needs significant work. They are constantly rechecking what the system is doing. The precision isn't where it should be. And as a computer scientist, again, these are solvable problems. Model accuracy improves with focus, better training data, and sustained engineering attention. But sustained engineering attention requires retaining the engineers. And retaining the engineers requires not telling them to go ask their parents about their career strategy when they request a day off. So, let's just take a tally for a moment. Meta has now attempted to be a social media company, yes. A metaverse company, $80 billion in losses. A frontier AI lab, $31.1 billion per quarter. A smart glasses company, glasses. A cloud provider, hinted at, no details yet, looking forward to it. A hardware company, demo failed on stage. An open source AI champion, gave away the models. A geopolitical weapon against China, sure. A personal super intelligence provider, as per the manifesto. And a baking assistant for the Zuckerberg family, also the manifesto. It has not, with the same energy, tried to be the best advertising AI company in the world, which is the one thing it is uniquely positioned to do, which is the thing that is about to overtake Google, which is the thing that generated $60.8 billion in a single quarter, which is the thing that funds everything else. Oh, and Meta is also now giving away its AI models for free to fight Chinese AI dominance. Forgot to mention that one. So, let's just add geopolitical benefactor distributing free technology to the world to the list. The list now has more items than the manifesto has concrete plans. Advertising is not something to be ashamed of. Imagine you started a business, you build something you believe in, you pour your savings into it, and nobody comes because nobody even knows you exist. Well, how would they? Advertising is how. It raises awareness, it creates markets, it connects people with things they didn't know they even needed. It is one of the most valuable functions in commerce, and Meta is the best at it in the world. Just do that. With great power comes great responsibility. Spider-Man's uncle told us that. The manifesto told us that Zuckerberg missed that lesson, likely because Spider-Man's uncle is sadly not Mark's uncle. What this manifesto really reveals is less a vision for the future and more actual evidence of a CEO who finds his own company's core business uninteresting. And the company whose CEO is bored by the thing that makes it exceptional is a company that will slowly and then suddenly stop being exceptional at it. Mark Zuckerberg is the ultimate trust fund child of Silicon Valley. A man who inherited the most powerful cash printing advertising engine in human history, got bored of it, and is now trying to burn down the casino just so he can pretend he invented fire. The diamonds are in the ground. The mine is right over there. And the CEO is writing about how one day diamonds will rain from the sky. If you want to understand how we actually got here, the superyacht, the morale crisis, the water contamination, and the $80 billion metaverse, I covered that across several videos. I'm going to leave links down below. Thank you for watching. I'll see you in the next one.