Behind Sam Altman’s ‘Code Red’ Response to OpenAI Competitors’ Profits

Behind Sam Altman’s ‘Code Red’ Response to OpenAI Competitors’ Profits
Behind Sam Altman’s ‘Code Red’ Response to OpenAI Competitors’ Profits

“I expect the vibes to be rough for a while.” That’s what Sam Altman said in a memo to staff last month. And the change in vibes has been really fast and furious.

The leader of OpenAI declared a “code red” to staff in a new memo, first reported by The Information, as the world’s most valuable startup is besieged by a wave of emerging AI rivals.

SUBSCRIBE: Receive more from our free The Daily Upside newsletter. READ ALSO: White House curbs fuel economy standards and Quiet $3 Trillion Investment Firm Set for Closing

Altman’s memo came out just as Google launched its impressive new Gemini 3. The Alphabet subsidiary said its new chatbot outperformed ChatGPT and Anthropic’s Claude in a series of key tests, and Altman admitted that the launch could “create some temporary economic headwinds” for OpenAI. Take it from a convert, Salesforce CEO Marc Benioff, who last month tweeted: “I’ve used ChatGPT every day for 3 years. I only spent 2 hours on Gemini 3. I’m not going back.”

ChatGPT still maintains a huge and growing weekly user base of 800 million people and is backed by world-class research from OpenAI. In this week’s memo, Altman promised the release of a new reasoning model next week that will surpass Gemini. Still, the number of Google chatbot users is increasing, up to 650 million in October from 450 million in July. The search giant has the built-in advantage of offering Gemini through integration with its other products. Calling it a “critical moment,” Altman will focus on improving ChatGPT at the expense of other projects including healthcare, purchasing, and personal assistant AI agents. A quick look at the financials makes it clear why it is imperative to shore up the company’s core product:

  • Anthropic, whose Claude has won commercial clients for its powerful encryption capabilities, forecasts it will break even in 2028, according to documents obtained last month by He Wall Street Journal. Meanwhile, OpenAI expects $74 billion in operating losses that year, equal to 75% of revenue, and doesn’t forecast a profit until 2030.

  • With $1.4 trillion in computing commitments through 2033, OpenAI needs continued growth to keep up with expected cash burn. HSBC analysts determined that even if its revenue reaches $200 billion by 2030 (up from about $20 billion this year), the company will need to cover a $207 billion shortfall with debt, equity fundraising or accelerated revenue growth.

It’s all connected: Google’s advances in artificial intelligence are causing headaches everywhere. The search giant trains Gemini on its own Tensor Processing Units (TPUs), rather than semiconductor giant Nvidia’s graphics processing units (GPUs), which are popular among its rivals. While Nvidia’s chips are more versatile, specially designed AI TPUs are cheaper and use less power, and Google’s AI boom has caused some nervousness among investors. Nvidia is down about 1.2% since the Gemini 3 was released on Nov. 18, while Broadcom, which helped develop Google’s chips, is up 11.5%.

This post first appeared on The Daily Upside. For sharp analysis and insights on all things finance, economics and markets, subscribe to our free newsletter The Daily Upside.

Source link