Power Shifts in Global AI

Nova Drake

Hey everyone, welcome back to the AI News Summary. I’m Nova Drake, here with Sam Guss. This week, the global chessboard just got a lot more interesting—let’s start in Shanghai, where Premier Li Qiang dropped a pretty bold proposal at the World AI Conference. He’s calling for a Global AI Cooperation Organization, basically a new international body to set the rules for AI. The idea is to counter what he calls fragmented rule-making and, let’s be real, the U.S. running the show. They want the headquarters in Shanghai and are really pushing to bring in Global South countries. It’s a big move.

Sam Guss - Avatar

It is, Nova. And it’s not just about governance on paper. This is China making a play to shape the very architecture of global AI—who gets a seat at the table, who sets the standards, and, ultimately, who benefits. There’s a kind of paradox here: the language is all about inclusivity, but the subtext is power. If you look at the last decade, we’ve seen the U.S. and Europe try to set the tone with their own frameworks. Now China’s saying, “Let’s do this together,” but on their terms. It’s a classic move in international politics—invite everyone, but keep the pen in your own hand.

Nova Drake

Yeah, and it’s not just talk. At the same conference, Huawei rolled out their CloudMatrix 384 system. This thing is a beast—384 of their own AI chips, all homegrown. Analysts are saying it’s right up there with Nvidia’s GB200, maybe even outpacing it in some tasks. I actually got to see some of this geopolitical maneuvering up close last year at a global tech expo. You could feel the tension—every booth, every handshake, it was like, “Who’s got the edge?” And now, with these hardware leaps, China’s not just talking about independence from U.S. chips—they’re building it, fast.

Sam Guss - Avatar

That’s the real story, isn’t it? The hardware race is the foundation for everything else. If you control the chips, you control the future of AI. Huawei’s move is a direct response to U.S. export controls, and it signals a shift from dependency to self-sufficiency. But there’s another layer: by pushing for a global governance body and simultaneously building out domestic infrastructure, China’s hedging its bets. If the international system doesn’t go their way, they’ve still got the hardware to run their own show. It’s a dual strategy—shape the rules, and make sure you can win no matter what the rules are.

Nova Drake

And the global South angle is interesting, too. China’s saying, “Hey, we’re not just building for ourselves, we’re building for everyone who’s been left out.” But until those countries have access to the same hardware and infrastructure, it’s more of an invitation than a reality. Still, it’s a narrative that’s gaining traction, especially as the U.S. and Europe double down on their own AI alliances.

Nova Drake

Alright, let’s jump across the Pacific. The U.S. just dropped its new AI Action Plan, and it’s basically a green light for the industry. We’re talking accelerated approvals, fewer state-level restrictions, and a clear message: let the market lead. It’s a big shift from the “let’s be careful” vibe we saw even a year ago.

Sam Guss - Avatar

It’s a classic American move—when in doubt, deregulate. But there’s a pattern here, Nova. If you look back at previous tech booms—think telecom in the ‘90s, or fintech in the 2010s—deregulation always comes with a promise of innovation, but it also brings hidden risks. The current plan is all about speed: fast-tracking data center approvals, limiting what states can restrict, and putting industry priorities front and center. But what gets lost in the rush? Oversight, safeguards, the slow work of making sure these systems don’t just work, but work for everyone. I might be wrong, but history says the pendulum always swings back when the cracks start to show.

Nova Drake

And then there’s the DOGE tool. This is wild—Musk’s team built an AI that’s already being used in federal agencies to cut regulations. Like, thousands of rules gone in days. The goal is to wipe out half of Washington’s red tape in a year. On paper, it’s efficiency. But, Sam, I keep thinking—what happens when an algorithm decides which rules matter? There’s no way it’s catching all the nuance, right?

Sam Guss - Avatar

That’s the ethical and legal gray zone, Nova. Automating deregulation is bold, but it’s also dangerous. Some of those “dead” rules are there for a reason—consumer protections, environmental checks, anti-bias measures. If you let an AI sweep them away without human review, you risk erasing hard-won safeguards. It’s the same story we’ve seen before: the promise of efficiency, but at what cost? And, honestly, the legal system isn’t built to keep up with this kind of speed. We’re rewriting the rules of governance, literally, and the consequences might not show up until it’s too late.

Nova Drake

It’s like we’re watching the regulatory process get eaten by its own tools. And the industry’s cheering, but I wonder how long before we see the backlash. I mean, we’ve talked about this before—every time there’s a wave of deregulation, it feels great until something breaks. Then everyone’s scrambling to put the guardrails back up.

Sam Guss - Avatar

Exactly. And the speed of AI just amplifies that cycle. The DOGE tool is a glimpse of what happens when you let the machines write—or erase—the rules. It’s efficient, but it’s also a warning. We need to ask: who’s really in control, and who’s accountable when things go sideways?

Nova Drake

Let’s shift gears to something a little more human—literally. The AtCoder World Tour Finals in Tokyo just wrapped, and Polish coder Przemysław Dębiak edged out OpenAI’s AI by