- In today’s CEO Daily: Diane Brady on what CEOs are saying about the AI bubble.
- The big story: Trump will provide Ukraine with air cover but not boots on the ground.
- The markets: AI bubble fears fuel tech selloff.
- Plus: All the news and watercooler chat from Fortune.
Good morning. Are we in an AI bubble? And if so, how should CEOs be thinking about that? None other than Sam Altman is opining that investors are “
overexcited about AI.” Yet, as my
colleague Sharon Goldman points out, Altman also says he expects to invest trillions of dollars in building out data centers in the coming years.
A man who compares his latest product launch to
the Death Star isn’t one who shies away from hyperbole. Altman has repeatedly said artificial general intelligence, the period at which an AI model performs as well or better than humans at most tasks, could
soon be upon us. Competitor Dario Amodei of Anthropic agrees. For some, that’s a signal to Buy Buy Buy.
Bubbles can take different forms. The dot-com bubble enveloped hundreds of startups that had people bidding for groceries and buying dog food from a sock puppet, few of whom made profits. This time around, the wealth of the many is being funneled to a few. More than a third of the S&P 500’s market cap
comes from the “Magnificent Seven”—
Alphabet,
Amazon,
Apple,
Meta,
Microsoft,
Nvidia, and
Tesla. That’s been true for some time and, unlike many dot-com disasters, these companies show real revenue growth. I hear more concerns about their market power, not market price. Meanwhile, companies like OpenAI and Anthropic are minting billionaires by the day without raising a penny in the public markets.
But Altman is onto something: Bursting bubbles are historically bad for business. Investors tend to be kind to players that are pumping up their portfolios. Attitudes change when markets crash. Disgruntled shareholders file securities class action suits, claiming they were deceived. Governments are pressured to step in to make sure the carnage never happens again. Heroes become villains and the spotlight shines brighter on issues that didn’t seem as urgent when times were good.
There’s also a nagging feeling growing among leaders I talk to that what’s great for the titans of tech might not be so great for the rest of us. In March, Altman
embraced Anthropic’s “Model Context Protocol” as an open-source standard for connecting AI models to external data sources and tools. Google
Deepmind CEO Demis Hassabis soon
embraced it, too.
Steve Lucas, the CEO of enterprise software company Boomi, told me yesterday that he’s suspicious of MCP, arguing, “there’s nothing that would stop a model from saying, ‘Oh, you have a nice MCP interface. Let me ask you how you work and reverse-engineer many of the functions of the application.’” That could cement their dominance, helping LLMs become the user interface for almost everything.
Contact CEO Daily via Diane Brady at diane.brady@fortune.com