No AI Good Guys? Andrew & Keith Ask If Altman Amodei, & Hegseth Have All Failed the Leadership Test

“They’re both naughty boys in the playground, leveraging the absence of clarity to their own advantage. Neither one of them is an authoritative leader of opinion with the interests of everyone at heart.” — Keith Teare

What a difference a week makes. Last Saturday, Keith Teare was arguing that Anthropic was wrong to push back against the US government’s use of AI in warfare. This week his editorial is entitled “No Good Guys.” He’s used AI to put images of Sam Altman, Dario Amodei, and Pete Hegseth around the same table—and found all three guilty of poor leadership. According to Keith, Amodei is “ideologically” (whatever that means) driven. Altman is commercially driven and Hegseth is just following orders. None of them is asking the all-important questions about AI policy. And the man who should be—Trump’s AI czar David Sacks—is absent-without-leave. All four should be court martialed.

Yes, a lot has happened in seven days. Altman publicly supported Amodei’s position on surveillance and autonomous weapons—then pulled a classic Sam u-turn and signed a contract with the Department of War. Amodei’s internal memo was leaked to The Information, revealing that he’d interpreted the government’s “no unlawful use” language as meaning there is no law. And the US military used Claude in the Iran war anyway. As Keith puts it: they’re all naughty boys in the playground, leveraging the gaps to their own self-advantage.

The only problem, of course, is that this isn’t a playground game. And that these men are all shaping the lives (and deaths) of countless people around the world.

Meanwhile, Om Malik’s “Post of the Week” offers a devastating contrast between Xi’s China and Trump’s America. China, Om argues, has published a five-year AI plan built on open-source software and bottom-up adoption. America, in contrast, has AI theater. No strategy, no policy, no leadership—just contracts, leaks, and perpetual spin. Then there’s the Startup of the Week, Jobright, which hit $5 million in annual revenue with nine people, suggesting that the companies of the future may not need humans at all. Keith’s own SignalRank has four people and claims to be going public. We seem to be heading for post-human companies before we’ve figured out who’s managing the humans.

Maybe we should court martial everyone. What a difference a week makes.

Five Takeaways

•       No Good Guys: Keith Teare’s editorial puts Sam Altman, Dario Amodei, and Pete Hegseth in the same room—and finds all three guilty of bad leadership. Amodei is ideologically driven, Altman is commercially driven, and Hegseth is just doing his job. None of them is asking the big questions about AI policy. The real culprit may be the invisible AI czar, David Sacks.

•       Altman Said One Thing, Then Did Another: Last week Altman publicly supported Amodei’s position on surveillance and autonomous weapons. This week he signed a contract with the Department of War. The contract uses “no unlawful use” language—which, as Amodei’s leaked memo points out, effectively means there is no law.

•       The US Used Claude in Iran Anyway: Despite the very public dispute between Anthropic and the government, the US military used Claude in the Iran operation. The government doesn’t need your permission to use your product. It just needs an API key and a credit card.

•       China Has a Plan. America Has Theater: Om Malik’s “Post of the Week” contrasts China’s published five-year AI strategy—built on open-source software and bottom-up adoption—with America’s complete absence of AI policy. The Chinese approach is more inclusive