The Wire Weekly
Episode 12 · July 6, 2026 · 00:08:44.45

This Week's Stories

Full Transcript

Strategy just gave itself permission to dump one point two five billion dollars in Bitcoin. The company formerly known as MicroStrategy. The firm that became synonymous with Bitcoin maximalism. The guy who wouldn't shut up about never selling. Yeah, that guy just rewrote his own playbook to include an exit door. And before you freak out, no, this isn't Michael Saylor going bearish. It's something way more interesting. It's him finally admitting that even the most diamond hands need liquidity engineering in a world where the rules keep changing. That's the world we're living in. I'm Travis Wright, this is The Wire Weekly, and this week the theme is simple. The moats everyone thought existed? They're evaporating. The competitors everyone dismissed? They're catching up. And the money everyone assumed would stay in Silicon Valley? It's flowing to places with actual sovereign balance sheets. Let's start with Strategy. The Bitcoin filing everyone's losing their minds over isn't a pivot. It's a hedge. They're not saying they're GOING to sell. They're saying they CAN sell up to a billion and a quarter if they need to manage their capital structure. Which, if you've been paying attention to how levered they are, makes total sense. This is Michael Saylor doing what he's always done. Playing four dimensional chess with debt, equity, and now apparently optionality. The man built a treasury strategy on never selling. Now he's adding a release valve. Why? Because the market's getting weird. Volatility is compressing. Institutional players are entering. And when you're sitting on that much BTC with that much borrowed capital, you need MORE tools, not fewer. This isn't weakness. It's sophistication. The people screaming betrayal don't understand corporate finance. The people nodding quietly? They get it. You don't become the biggest corporate Bitcoin holder by being inflexible. You do it by being smarter than everyone else in the room. And right now, the smart move is keeping your options open while the macro picture sorts itself out. But here's where it gets REALLY interesting. While Saylor's building flexibility into his balance sheet, the rest of the world is realizing that the competitive advantages they thought were permanent... aren't. Meta just admitted internally that their AI agents are flopping. Zuckerberg told his own staff that progress is slower than expected. Tens of billions of dollars in compute. Some of the best researchers on the planet. And they're BEHIND schedule. Meanwhile, Meta's Llama 4 model, according to Scale AI's chief, now matches GPT-4o. Not close. Not almost. MATCHES. Which means the gap between OpenAI's best and everyone else's best just collapsed to basically zero. And if you're OpenAI, that's a nightmare scenario. Because your whole valuation, your whole narrative, your whole moat was predicated on being six to twelve months ahead. If Meta can catch you with open source models that anyone can download and run, what exactly are people paying you for? Access? Convenience? That's a different business. That's not a hundred billion dollar win the future of intelligence business. That's a SaaS company with really good APIs. And it's not just Meta closing the gap. Mistral AI in Europe just hit six billion in valuation. You can download their models. You can run them locally. You can fine tune them on your own data. They're not trying to lock you into some walled garden. They're building the open source alternative to everything OpenAI represents. And the market is REWARDING them for it. Meanwhile, DeepSeek's new coding tool costs ninety seven percent less than GitHub Copilot. Not a typo. Ninety seven percent. Same functionality. Fraction of the price. And American developers are about to become the beneficiaries of what is effectively a geopolitical subsidy race. Because when Chinese AI labs are competing with American AI labs, the people who win are the customers. The moats aren't just shrinking. They're being commoditized in real time. Now let's talk about the Trump administration handing out access to Anthropic's Mythos model like it's candy on Halloween. Anthropic built this thing. They tested it. They realized it was so good at finding software vulnerabilities that they locked it away. Too dangerous. Not ready for prime time. Then the Trump administration overruled them and gave access to over a hundred organizations, including their foreign employees. Which is either the most reckless thing I've heard this year or a calculated geopolitical flex. Maybe both. If you're Anthropic, you just got steamrolled by your own government. If you're a security researcher, you're either thrilled or terrified depending on which side of the vulnerability marketplace you're on. And if you're China or Russia, you're taking notes. Because this is what happens when national security policy collides with AI safety norms. Someone's priorities win. And right now, it's not the safety researchers. But you know what IS working? Crusoe Energy. These guys are raising three billion dollars at a nine billion dollar valuation. And before you roll your eyes at another overvalued AI infrastructure play, hear me out. They're not building data centers in Northern Virginia. They're taking stranded natural gas, the stuff that gets flared off at oil wells because there's no pipeline to move it, and they're using it to power compute. AI training. Bitcoin mining. Cloud workloads. Energy that was being wasted is now being monetized. And the market just said that's worth nine billion dollars. This is the infrastructure layer of what I'm starting to call Web4. Not crypto. Not just AI. The convergence of compute, energy, and economic incentives in places where traditional players can't or won't go. If you're paying attention, this is where the next decade gets built. Not in Silicon Valley. In the middle of nowhere, next to a gas flare, with GPUs humming and capital flowing from people who actually understand energy arbitrage. Which brings me to the actual story of the week. Abu Dhabi. MGX, a two year old fund, just raised forty nine billion dollars for AI investments. Not million. Billion. With a B. That's more than every US venture capital fund COMBINED raised last quarter. They're not investing in AI companies. They're buying the ENTIRE stack. Compute. Data centers. Model development. Applications. The whole thing. And they're doing it with sovereign wealth money, which means they don't need exits on a Sand Hill Road timeline. They can play the long game. They can subsidize. They can out wait everyone. While Silicon Valley is fighting over who gets the next Sequoia term sheet, Abu Dhabi is writing checks that make Sequoia look like a seed fund. This is the petrostate playbook. They did it with real estate. They did it with sports. Now they're doing it with AI. And if you think they're just dumb money chasing hype, you're not paying attention. These guys hired the smartest people in finance, pointed them at the future, and said go buy it. And that's exactly what they're doing. So what should you be watching? Three things. One, watch how OpenAI responds to the commoditization pressure. If everyone can match your models in six months, you either need to innovate faster or you need to pivot your business model. I think they're going to do both. Two, watch European AI companies. Mistral isn't alone. There's a whole wave of open source, privacy forward, non US AI development happening. And as trust in American tech continues to erode globally, that's going to matter. Three, watch the Middle East. Not just the money. The STRATEGY. They're not trying to copy Silicon Valley. They're building something different. And if they succeed, the next decade of tech won't be centered in California. It'll be distributed. Multi polar. And a hell of a lot more competitive. The moats are gone. The competition is global. And the money is flowing to whoever has the longest time horizon and the deepest pockets. That's not the world we thought we were building five years ago. But it's the one we're living in now. I'm Travis Wright. See you next week.