Grading the CHIPS Act and the other costs of AI

Hello from sunny California. I'm in Los Angeles this week, along with thousands of bankers, investors, policymakers and tech executives, for the annual Milken Institute Global Conference.

While the conference, now in its 27th year, is known as an exclusive event for financiers, tech heavyweights are increasingly drawn to the gathering in palm tree-lined Beverly Hills. For example, Elon Musk skipped the Met Gala this year and joined the conference on Monday evening in a session where he shared his visions for AI and colonizing Mars, as well as his concerns about declining populations and illegal immigration.

Musk was not the only one talking about AI at the conference. The words "artificial intelligence" came up at least five times in every panel and sideline meeting I attended, and even in casual conversations with fellow conference attendees while waiting for our Ubers.

While the theme of this year's conference is "Shaping a Shared Future," much of the discussion about AI centered on how the U.S. should compete with China on this technology. A panel on the geopolitics of AI said there was "not a lot" of hope to include Beijing in the talks about AI safety and regulation that the U.S. is having with its allies. Even the idea of the U.N. serving as a potential platform for building a global AI governance framework was dismissed.

"The more you do things in the U.N., the more you play into China and Russia's hands," said Karen Pierce, Britain's Ambassador to the U.S.

I left the Milken conference wondering: How can we govern this immensely powerful and rapidly developing technology when there is such a slim chance that the world's two AI superpowers can work together on this issue?

Artificial intelligence is costly, and not only