The number is so large it breaks your brain. They raised more money than some countries' GDP and still managed to lose Disney.
$120 billion. Not million. Not "over a hundred million." One hundred and twenty billion dollars — the largest private capital raise in the history of technology. OpenAI closed that round this week. And in the same seven days, they killed their flagship video product and lost a $1 billion Disney partnership. Normal week. ---
Let's actually sit with the number for a second.
In February, OpenAI raised $110 billion at a $730 billion valuation. Amazon came in at $50B. SoftBank $30B. Nvidia $30B. Already the biggest private tech raise ever. Then this Monday, Bloomberg broke the news: OpenAI was bolting on another $10 billion from MGX, Coatue, Thrive, and Altimeter. Total: $120+ billion. To put that in perspective — that valuation makes OpenAI bigger than Netflix, Goldman Sachs, Nike, Starbucks, and Ford combined. Now here's where it gets genuinely strange. In the same week they closed this earth-shattering raise, OpenAI killed Sora.
Yes — the AI video generation tool launched with massive fanfare in 2024, the one that was supposedly going to revolutionize creative content forever. Shut down. And with it? A $1 billion licensing deal with Disney for AI-generated content. Disney walked. The Hollywood Reporter confirmed it Wednesday. A billion-dollar partnership — gone. This is wild for a few reasons. One: OpenAI is so awash in capital that losing $1B barely registers. Two: the companies who built workflows and business models on top of Sora just got the rug pulled. Three: it proves that in 2026, no AI product is too big to kill. Even the flagship ones with household-name clients get discontinued when the strategy shifts.
The lesson isn't "don't use AI tools." It's: never build a single-point-of-failure dependency on any one platform. The companies that are winning right now are the ones who treat AI tools like utilities — interchangeable, pluggable, never the foundation. The speed of all this is genuinely hard to track. Build fast. Stay diversified. Don't let anyone's product roadmap become your business model. ---
3 OTHER WTF MOMENTS THIS WEEK
Meta fired hundreds of employees (Reality Labs, recruiting, sales) while simultaneously announcing "exceedingly aggressive" stock bonuses for top executives. Cut the workers, reward the bosses, pour everything into AI infrastructure. The playbook is now shameless. At least they're consistent.
AI agents are getting credit cards. Visa completed hundreds of real-world autonomous agent transactions this week and is targeting millions of AI-driven shoppers by the 2026 holiday season. Your AI assistant will soon be able to research, compare, and purchase things without you touching the checkout. That's either the most useful thing ever invented or a personal finance catastrophe waiting to happen.
Goldman Sachs quietly admitted this week that despite $660 billion in AI investment, there's been essentially zero measurable return on US GDP. Hundreds of billions in. Economy barely moved. The investment thesis is "it'll compound later." Maybe. But that Futurism piece quoting their analysis is the most important thing nobody talked about this week.
THE MONEY ANGLE Here's what nobody says out loud: every time a major AI product shuts down (Sora, now joining a long list), it creates opportunity. The people who built Sora-based workflows now need something new.
The businesses that relied on it need someone to help them migrate and rebuild. Meanwhile, that Goldman survey found 93% of small business owners using AI are seeing real results — but most of them have no idea how to actually implement it. That's a service gap. "AI setup and workflow consulting for small business" is a legitimate, growing income stream right now.
You don't need to raise $120 billion. You need a Stripe account, a strong prompt library, and the ability to get results for someone who doesn't know how.
→ Want the exact prompts powering real income right now? Grab the Prompt to Profit starter kit and start this week — not someday.