Three years ago, ChatGPT lit a fuse. Not just under OpenAI, but under the entire tech industry. Like the iPhone in 2007 or the launch of Skype in 2003, it wasn’t just a product drop — it was a moment. In the span of 36 months, the world moved from wondering what this chatbot thing was to seeing generative AI reshape how companies operate, invest, and even think.
AI went from science project to boardroom agenda. Enterprises like Walmart, BNY Mellon, and Intuit aren’t playing with pilot projects anymore. They’re deploying real AI in meaningful ways, not just talking about it at conferences. This isn’t hype. It’s execution.
That’s not to say there isn’t noise. There’s plenty of it. The classic Silicon Valley playbook, “fake it till you make it,” is crashing into the hard economics of generative AI. Running models isn’t cheap, and many startups are bleeding money per user. Growth at all costs doesn’t cut it when investors are asking: Where’s the revenue?
Just look at the bond market. The big AI backers like Amazon, Meta, Google, and Oracle have flooded it with $90B in debt to fund infrastructure and data centers. Speculative-grade bonds are coming from the likes of CoreWeave and TeraWulf, and yields are climbing. That kind of capital strain tells us this isn’t a normal tech cycle. This is a full-blown infrastructure transformation, more like telecom buildouts of the early 2000s than your usual SaaS boom.
Then there’s the product side. Google’s Gemini 3 leapfrogging ChatGPT in benchmarks and real-world use cases isn’t just a technical win. It’s a branding win. It signals that the AI “race” isn’t over. It’s accelerating.
Meanwhile, Meta’s accounting gymnastics—shifting a $27 billion data center off balance sheet into a joint venture—is the kind of financial footwork that feels ripped from a late-stage SoftBank portfolio play. Clever? Yes. Risky? Also yes.
Still, despite the shakiness, AI’s core remains resilient. We’re watching new business models form around embedded AI, with ChatGPT becoming a distribution channel in its own right. Companies want to be in it, not just using it.
The AI hype may be deflating, but that’s healthy. It’s making room for the builders, the real users, and those who see AI not just as a product but as a platform.
As I’ve written before, technology cycles start with buzz. Then come the bumps. What matters most is who builds through them .