Axios

We scrambled the debut of our "Behind the Curtain" video series: Anthropic CEO Dario Amodei wanted to chat Monday after he dropped a 38-page essay warning of escalating risks from the AI technology he's helping pioneer.

• We asked Amodei, in San Francisco, what Congress should do now, and what lawmakers should tell their constituents. His three-part prescription:

1. Transparency legislation to require AI companies to disclose their models' risks and bad behavior, and the defenses that are built in.
2. Cut off sales of Nvidia chips and other U.S. products that help China.
3. Get ready to tax future AI trillionaires and redistribute wealth. He said he'd tell his fellow future trillionaires: "You're going to get a mob coming for you if you don't do this in the right way."

The bottom line: "We always assume that everything that can go wrong does go wrong," Amodei said. "That's how you build things that are reliable."

👉 Share the YouTube ... Executive producer: Axios' Jimmy Shelton.

DAVOS, Switzerland — Widespread access to benefits and safety standards around AI is a collective responsibility, said industry leaders at a Jan. 21 Axios House event.

Why it matters: When the public feels the negative consequences but doesn't reap the rewards, "that will be an organ rejection when it comes to embracing the technology," TIAA president and CEO Thasunda Brown Duckett said.

Axios' Courtenay Brown and Ina Fried spoke with Brown Duckett, Google DeepMind co-founder and CEO Demis Hassabis, IBM vice chairman Gary Cohn and U.S. Trade Representative Jamieson Greer. The conversation was sponsored by ServiceNow.

Zoom out: Concerns about equity apply not only to financial benefits but also to resources and community impact.

• "We're starting to see the issue of water. We're starting to see the issue of electricity prices. These are real fundamental issues," Cohn told Brown.
• "We need to get our act together quickly on things like international cooperation, at least some kind of minimal safety standards," Hassabis said.

Zoom in: Preventing AI-exacerbated inequality "starts within your own company," Brown Duckett told Brown.

• "Are you taking a look at your wages? Are they competitive? Are you still recruiting on campuses?"
• "We have to make sure: How will everyone be able to have an opportunity to benefit?"

The bottom line: "There's a bigger picture at stake of safety overall and stewarding AGI safely into the world for the benefit of everyone," Hassabis said.

Content from the sponsor's segment:

In a View from The Top conversation, ServiceNow.org president and ServiceNow chief corporate affairs officer Vanessa Smith warned that failing to address inequality early risks repeating past mistakes: "History tells us what happens when those divides widen."

Go deeper: Watch the full interviews on YouTube.

Some of California's most prominent venture capitalists are quick to slam their state, arguing that fiscal mismanagement and high taxes will cause startups to form elsewhere.

• So far that doesn't seem to be happening.