Show Notes
We get into the love/hate triangle between Anthropic, OpenAI, and the Pentagon and discuss the details of these contracts for AI weapon and surveillance systems, what’s actually at stake here with debates over the terms of “guardrails”, “red lines” and “lawful uses,” and how the competing (a)moral visions for AI and war that are represented in this debate really come down to a difference of opinion about technical capabilities at this current moment, not about the fundamental ethics and politics at play in weaponized AI. Plus, why this lovers spat between Anthropic and Pentagon must not trick you into labelling Dario Amodei as a righteous purveyor of resistance tech. Get real!
••• How Talks Between Anthropic and the Defense Dept. Fell Apart https://www.nytimes.com/2026/03/01/technology/anthropic-defense-dept-openai-talks.html ••• Inside Anthropic’s Killer-Robot Dispute With the Pentagon https://www.theatlantic.com/technology/2026/03/inside-anthropics-killer-robot-dispute-with-the-pentagon/686200/ ••• The Adolescence of Technology https://www.darioamodei.com/essay/the-adolescence-of-technology ••• A Few Observations on AI Companies and Their Military Usage Policies https://sarahshoker.substack.com/p/a-few-observations-on-ai-companies ••• "All Lawful Use": Much More Than You Wanted To Know https://www.astralcodexten.com/p/all-lawful-use-much-more-than-you ••• Who loses from the Anthropic fight? Maybe Elon Musk and Alex Karp. https://www.programmablemutter.com/p/who-loses-from-the-anthropic-fight ••• AI vs. the pentagon https://jasmi.news/p/ai-pentagon ••• What Rights Do AI Companies Have in Government Contracts? https://jessicatillipman.com/what-rights-do-ai-companies-have-in-government-contracts/
Standing Plugs: ••• Order Jathan’s book: https://www.ucpress.edu/book/9780520398078/the-mechanic-and-the-luddite ••• Subscribe to Ed’s substack: https://substack.com/@thetechbubble ••• Subscribe to TMK on patreon for premium episodes: https://www.patreon.com/thismachinekills
Hosted by Jathan Sadowski (bsky.app/profile/jathansadowski.com) and Edward Ongweso Jr. (www.x.com/bigblackjacobin). Production / Music by Jereme Brown (bsky.app/profile/jebr.bsky.social)
More from This Machine Kills
We discuss and riff on the People’s AI Assembly, organized by AI Now Institute and More Perfect Union, which just happened in New York City.
Jathan chats with journalist Kate Kaye about her new podcast series, Trust Control, all about the history and future of trying to measure trust. We get into different ways of defining trust, proxies for turning trust into quantified scores and “objective” models, motivations for making trust into a metric that can be optimised, and ultimately, the anti-social shift from trust as a dynamic social relationship that is a core part of human interactions to trust as an objective feature of systems with or without humans.
We discuss a deep dive into the ascendency of microdramas – the short-form video series that are already bigger than the conventional film industry in China and are quickly taking over studios in Hollywood. We chat about the financial and cultural antecedents of microdramas, and the strange impacts of these videos on media production and consumption today, and the unevenly distributed futures of AI generated entertainment.
We get into the new decree from Trump that “artificial intelligence” is now called “super intelligence” – it sounds less fake and more powerful — and the broader currents of tech advice and policy in the Trump administration. Then we discuss how OpenAI and Anthropic are behaving like rogue agents, who are also irresponsible and incompetent to a degree that should be unbelievable.
We learn all about how leftists are against any new regulations for AI. It’s the unbelievable return of horseshoe theory as we discuss the curious new discourse cycle being promoted by prominent pundits.