OpenAI’s Explosive Growth Continues

I’m Tim Stenovec, in for Ed Ludlow. We’re on OpenAI’s reported route to $40 billion in annualized revenue, the AI spending race, and New York’s push to bring Amazon delivery workers in-house.

Consumer is still OpenAI’s biggest business after topping one billion weekly active users, but business adoption is climbing, including through Codex. Anthropic is a serious coding rival, yet OpenAI’s revenue shows customers are paying too.

The expense picture remains hazy, but compute costs a lot. OpenAI has built major capacity and still calls access to enough compute its biggest constraint, so it will consume a huge share of revenue.

Quarterly numbers may be the wrong lens. Hyperscalers and chip companies are planning through 2027, 2028, and 2029, borrowing for decades to fund spending against revenue promised years out.

Nobody knows the timing. Hyperscalers are in an existential race for future compute through data centers, Nvidia chips, and their own silicon. Returns should come, but compute needs are still not fully priced in.

Watch the familiar public names, where spending, borrowing, and circular financing sit. Nvidia faces customers building chips, CoreWeave borrows against Nvidia hardware, and concentration among giant buyers is real.

Applied Materials posted clean numbers, and investors said, “Great, we knew that. What’s next?” Its CEO sees eight quarters ahead because equipment takes so long to build and install. Next year improves further.

More than 80 percent of people in China tell Stanford they’re optimistic about AI, versus under 40 percent in the U.S. Chinese technology has meant empowerment; Americans often see addiction, privacy abuse, and concentrated power. That could give China an adoption edge.

Robotics has not had its ChatGPT moment, but you can see the light. We’re building an intelligence layer that understands new settings, reasons through them, and acts on unfamiliar tasks. Humanoids matter, but not for every job.

The delivery-service model keeps companies distant from workers and liability. Amazon controls uniforms, routes, hiring, and hours, yet can avoid accountability for unsafe productivity. Direct responsibility, licensing, and enforcement could protect workers, consumers, and city streets.

AI investing is noisy, so I want non-consensus founder stories that grow compelling as assumptions prove out. The future of work is bigger than chatbots: AI can perform tasks with human supervision. Jobs will shift, demanding reskilling.

We closed with a reminder that a giant audience does not make every creator expansion easy. Alex Cooper’s beverage brand is reportedly winding down, even as Unwell has a $500 million valuation. Have a great weekend.