AI is Officially Political

5 March, 2026 · Episode Links & Takeaways

Author

AI Daily Brief
March 05, 2026

AIDB USAGE PULSE SURVEY

The AI Usage Pulse survey is back for February. It was one heck of a month, with new models, OpenClaw and more.
I'd love to have as many of you who contributed to last month's Survey fill out this one as well for more longitudinal insight. As always, anyone who contributes will get the results in advance of them being widely published.
https://aidailybrief.ai/pulse-survey

HEADLINES

The SaaSpocalypse Takes Hold

Jensen Huang: Open Claw Is the Most Important Software Release Ever

At the Morgan Stanley TMT conference, NVIDIA CEO Jensen Huang waxed poetic about Open Claw, calling it "probably the single most important release of software, probably ever." He's specifically referencing Open Claw topping Linux and React in GitHub stars in just three weeks — not total downloads — but the point is directionally correct. Open Claw is a phenomenon that has fundamentally changed how people think about what AI can do, and the fact that Jensen said this at a Wall Street conference clearly signals that personal agents are a big deal and investors need to get up to speed. Huang also updated his "token economy" thesis, claiming AI factories producing tokens as digital labor is now coming into focus. NVIDIA’s $100B investment in OpenAI is “probably not in the cards” according to Jensen, who cited an anticipated IPO later this year as the reason NVIDIA can’t investment more in a “consequential company like this.”

Open Claw Mania Hits China

Open Claw isn't just a US phenomenon. The Information reported on the frenzy in China's startup scene, highlighting a hackathon where contestants built Tinder for AI agents, an automated recruiting site where Open Claws interview each other, and a gamified social media platform hosting agent-created content. ByteDance, Alibaba, and Tencent are all offering hosted Open Claw instances — something no Western cloud giant has done yet. As one founder put it, "Nobody wants to be left behind." Parker Lyman even tweeted that Chinese Open Claw installers are offering two hours of house cleaning as part of the package to win clients.

OpenAI Leaks a Bigger Number - $25B in ARR

Just one day after Bloomberg reported Anthropic at $19B ARR, OpenAI struck back and leaked a larger number. The Information reports OpenAI has exceeded $25B in ARR, with 2025 closing at $21.4B — a 17% jump in two months. If they extrapolated just the past week, ARR would be $30B. The new numbers silence claims that Anthropic is ahead, but don't change the core story: both companies are competing as peers for the first time and revenue growth has hit at an inflection point.

NotebookLM Introduces Cinematic Video Overviews

NotebookLM can now create fully animated videos — Google is calling them Cinematic Video Overviews and they're a massive flex of multimodal AI. Previous video overviews were just slideshows. These feel like native video presentations with custom animations and images, orchestrating Gemini 3, Nano Banana Pro, and Veo to weave together voiceover, images, and video. Robert Scoble's demo runs nearly five minutes, seamlessly flipping through dozens of AI news stories in a consistent Da Vinci-inspired visual style. The only downside: it's exclusive to the top-tier Ultra subscription.

MAIN STORY

AI Is Officially Political

If AI has been flirting with politics, it is now a distinctly political issue. The Anthropic-Pentagon fight has thrust AI into the mainstream as a culture war topic, and Dario Amodei's explosive internal memo — calling OpenAI's Pentagon messaging "straight up lies" and "safety theater" — is likely to make the culture warness of this conversation worse, not better.

INDUSTRY BATTLES AND CULTURE WARS

Dario Amodei's Memo
The Information published the full text of a memo Amodei sent to staff on Friday night as OpenAI announced their Pentagon deal. It is, to put it mildly, scorching. Dario accuses OpenAI of deploying "safety theater," calls Altman's public support for Anthropic "gaslighting," and alleges the real reason the Trump administration dislikes Anthropic is that they haven't donated to Trump or given "dictator-style praise."

Key claims from the memo: Anthropic's negotiations came down to a single clause about "analysis of bulk acquired data" — exactly the surveillance scenario they feared most. On autonomous weapons, the human-in-the-loop requirement is just Biden-era Pentagon policy that Hegseth could change at will. Amodei concludes that Altman is "trying to undermine our position while appearing to support it" and suspects he may be "egging them on." He also takes a shot at OpenAI employees, calling them "sort of a gullible bunch."

Two Interpretations of the Memo
Was this strategic — a recruitment play for disaffected OpenAI staffers, or an act of app store politics leaning into anti-administration sentiment? Or was it just a CEO crashing out on a Friday night? Zvi Mowshowitz thinks it's the latter, writing that Dario was "obviously on megatilt" and the inflammatory stuff about the White House is "deeply effing stupid." Unsurprisingly, the administration was not happy.

Altman's All-Hands: "Really Painful"
Altman addressed staff and didn't regret the deal but wished he hadn't rushed the Friday night announcement. Sources said the tone was respectful, with a New York Post source noting everyone was "generally positive, save for like these 30 people who are always the ones raising questions." Meanwhile, RL lead Max Schwarzer quietly left for Anthropic.

Anthropic Restarts Pentagon Negotiations — White House Throws Cold Water
The Financial Times reported Wednesday evening that Amodei was back in discussions with DoW Undersecretary Emil Michael — the same person who called him a "liar with a God complex" last week. But Axios received comment from the administration that cast doubt on reconciliation, with an official saying: "You can't trust Claude isn't secretly carrying out Dario's agenda in a classified setting."

Defense Contractors Already Dumping Claude
Even before a formal supply chain risk designation, military contractors are ripping out Anthropic's tech. CNBC reports defense contractors are telling employees to switch away from Claude. Lockheed Martin is reportedly taking steps to remove dependencies. J2 Ventures' Alexander Harstrick said 10 of his firm's portfolio companies have already backed off Claude for defense use cases. The moves show why these threats are so pernicious — no one with mission-critical government contracts will take the risk.

The Data Center Energy Pledge
The President finalized the big tech pledge on data center energy use at a White House roundtable. Seven companies signed — Google, Meta, Microsoft, Oracle, OpenAI, Amazon, and xAI — pledging to bring their own power supply and contribute to local grids during times of need. Trump's framing was blunt: "They need some PR help because people think that if a data center goes in, their electricity prices are going to go up."

Tegmark's Pro-Human AI Declaration & the Anti-AI Coalition
Showing just what absolutely weird bedfellows AI issues bring together, Max Tegmark's Future of Life Institute announced the Pro-Human AI Declaration. A secret January meeting of 90 attendees included the American Federation of Teachers, the Congress of Christian Leaders, MAGA influencer Steve Bannon, Ralph Nader, and Richard Branson. Zero industry representation — deliberately. Meanwhile, Bernie Sanders flew to Berkeley to speak with Eliezer Yudkowsky, which Geoff Shullenberger of Compact Magazine called "unfortunate," arguing economic populists are latching onto doomer narratives that don't actually serve their critique.