
Satya Nadella says he’s on AGI pill now
Microsoft CEO Satya Nadella says he’s a fan of AGI, at least when it comes to how AI could bring surprising changes to the structure of the economy and how policy might change as a result. “I am AGI-pilled in the sense that at the economic level, what is the organization of the political economy around AGI?” he asked during a panel discussion after his Microsoft Ignite keynote at the Chase Center in San Francisco.
But Nadella doesn’t think pioneering models like Claude are aware. “I don’t think so,” he said, adding that he followed Microsoft AI CEO Mustafa Suleyman’s view that companies should not grant rights to models and treat them as if they could have moral status, which is what companies like Anthropic do on principle. Suleyman told me he thought the Anthropic team had been “stubborn” and that this approach was dangerous.
Nadella wasn’t as firm about it. “It’s a good debate,” he said. “I’m not dogmatic about this at all. What I’m saying is that the most important social technical problem we face is how do we ensure that an economy can disseminate this technology on a large scale?”
I asked Nadella what he thought about AI after hearing his response to another reporter’s questions about Anthropic launching more products in its office suite. Nadella has spoken at length about the potential collapse of the economy and the end of democracy if an AI model could achieve monopoly status.
“If the Marxist view is that there is only one company, then it’s over, right? For all of us.”
“Most people see this as a competition between technology companies, which of course it will be,” he said. “But remember, the model has no limits. It will be a health care company. It can be a financial services company. It can be anything if it wants. It can take over the government. Thinking of this as a close competition among technology companies, to determine who gets the next trillion dollars in market cap, is one of the strangest, narrowest ways to think about it.”
Nadella added that the very nature of business risks being disrupted in the future. “If intelligence has network effects, what is the future of business? he asked. “If the model sees something and learns it, and virtually all human capital collapses into symbolic capital, what is the role of a business and human capital in a business? That’s the debate.”
“I think political economy is going to become more and more interesting in this context. Nobody – at least in a democracy – can even win an election if things are like this, if there’s only capital and no labor. I mean, I don’t know, it’s not an economy.”
“If the Marxist point of view is that there’s only one company, then it’s over, right? For all of us. Not just for me. You’re also, in one way or another, going to be a subsidiary of Anthropic.”
Needless to say, that’s not how tech CEOs usually talk about the launch of competing spreadsheet software.
Of course, Nadella was also promoting Microsoft’s various AI tools at Ignite, which promise to allow enterprise customers to choose the most cost-effective models for various tasks, run their own models, and control their own data in different ways. The idea here is that the market will demand many different types of models and sovereign data in order to prevent centralization of control, a theme that Nadella kept returning to as we spoke.
“I think if you think the global economy needs more than one or two companies, then by definition there have to be as many models as there are companies,” he said.
Gn tech