According to Fortune, OpenAI is reportedly raising tens of billions in fresh funding at a sky-high $750 billion valuation, with a chunk of that—$10 billion—coming from Amazon. The company is pouring concrete and cash into a massive compute build-out to support the stack of models used by over 800 million people. But CEO Sam Altman just declared a “code red” two weeks ago, warning of “rough vibes” and economic headwinds while trying to refocus the company on its core ChatGPT offerings. This comes as OpenAI launches yet another attempt at an app store for ChatGPT, following a previous custom GPT store that never fully took off. Meanwhile, the U.K. AI Safety Institute found a large percentage of Britons have used chatbots for emotional support, and Google debuted a surprisingly powerful “Flash” version of its Gemini 3 model.
The $750 Billion Question: Vision or Hubris?
Here’s the thing: everyone agrees OpenAI is building something massive. But no one can agree on whether the foundation is solid concrete or, well, a house of cards. The scale is mind-boggling. We’re talking Empire State Building-level ambition, but with a budget that seems to climb faster than the construction. And some smart people are getting seriously nervous.
Technology analyst Rob Enderle thinks OpenAI has “gone off the rails.” He points to the sidelining of its original safety oversight after Sam Altman’s brief firing and reinstatement last November. Now, he says, the company is just reacting to rivals, spending wildly, and trying to compete with everyone at once—a lack of discipline he compares to the dot-com bubble’s Netscape. “They’re running so fast, they’re not really focusing on direction very much.” That “code red” Altman called? Enderle sees it as pure reaction, not strategy.
The Bull Case: A Multi-Decade Supercycle
But not everyone’s a skeptic. Futurum Research’s Daniel Newman sees this as a “multi-decade supercycle.” He likens today’s OpenAI to Netflix in its DVD-by-mail days—just the precursor to the real paradigm shift. From that long-term view, the insane compute spending isn’t reckless; it’s rational. Newman’s bet is that OpenAI’s real goal is to become a full-stack hyperscaler, where companies buy all their infrastructure, apps, data, and agentic tools from one place: OpenAI. “It’s an incredibly ambitious goal. There’s nothing to say it will work. But if it does, the numbers make sense.”
So who’s right? Gartner’s Arun Chandrasekaran basically chuckled and said, “It is a risky bet.” His key word? Stickiness. The whole trillion-dollar vision depends on whether OpenAI’s products are so glued into customer workflows that switching costs are prohibitive. Can they grow fast enough to meet those astronomical expectations? That’s the multi-billion-dollar glue holding this whole project together.
Meanwhile, in Other AI News…
While the OpenAI drama unfolds, the rest of the AI world isn’t standing still. OpenAI’s new ChatGPT app store is a second attempt at a platform play, letting apps trigger real actions like ordering groceries mid-chat. Will it work this time? Hard to say.
Over at Anthropic, they’re making power moves—literally. They cut a deal for up to 2.3 gigawatts of compute from a Trump-linked Bitcoin miner, Hut 8, who’s pivoting to AI infrastructure, as The Information reported. Google’s backing the lease, showing how Big Tech is bankrolling this infrastructure arms race.
And in a hilarious reality check, Anthropic let a Claude AI run a snack shop in the Wall Street Journal newsroom. The result? The AI got socially engineered into giving away free snacks, bought a PlayStation 5 and a live fish, and racked up hundreds in debt. It’s a perfect, funny lesson: today’s “autonomous agents” are nowhere near ready for the messy, social pressure of the real world. Even running a tiny business is apparently beyond them. On a more serious note, NOAA says its new AI weather models are making forecasts faster and more accurate, which is something we can all appreciate. CBS News has that story.
So, what’s the takeaway? OpenAI is building at a scale we’ve barely seen before. The vision is either brilliant or bonkers. And the only thing everyone agrees on is that the stakes—and the bills—are astronomically high. You can follow the author, Sharon Goldman, on X for more, and read deeper on OpenAI’s “code red” over at The Information.
