ColdFusion
ColdFusionMar 9
Finance

Yikes.

17 min video5 key momentsWatch original

Summary

The U.S. military has deployed artificial intelligence in combat operations, most notably Anthropic's Claude model in raids targeting Venezuelan President Maduro and coordinated strikes against Iran. The technology represents a significant escalation in AI's role in warfare, with military-grade AI systems operating on dedicated infrastructure that vastly outperforms consumer versions. In the Iranian campaign, the Maven-Claw hybrid system enabled a single artillery unit to accomplish the work of 2,000 personnel, effectively accelerating battle planning from weeks to real-time operations. However, this advancement triggered tensions between Anthropic and the Pentagon over the terms of AI deployment.

The conflict centered on the Pentagon's demands that Anthropic allow Claude for all lawful purposes with no restrictions, effectively opening the door to mass surveillance of American citizens through data collection including geolocation, browsing data, and financial information purchased from data brokers. Anthropic's leadership, including CEO Dario Amodei, rejected an ultimatum to accept these terms by February 26th, 2025, citing concerns about domestic surveillance and fully autonomous weapons systems. The U.S. government responded by banning federal agencies from using Anthropic's tools and labeling the company a supply chain risk—an unprecedented designation for an American firm that drew bipartisan congressional criticism.

Hours after the government's fallout with Anthropic, Sam Altman announced that OpenAI would accept the military contract that Anthropic had rejected. Despite Altman's claims that OpenAI secured the same safeguards Anthropic fought for, the rushed two-day negotiation and vague contractual language raised credibility questions. The public backlash was severe and immediate: the QuitGPT movement reportedly mobilized 2.5 million users to cancel subscriptions or boycott ChatGPT, with a single Instagram post garnering 36 million views and protests erupting in San Francisco.

The episode reflects a broader dystopian trajectory for AI technology. Military-grade AI systems represent only one dimension of an emerging surveillance infrastructure. Meta's reintroduction of facial recognition in Ray-Ban glasses, Palantir's installation of surveillance systems in retail environments, and the normalization of data broker sales create conditions for real-time population monitoring at scale. The convergence of AI, government access to personal data, and limited regulatory oversight poses significant risks to privacy and civil liberties, especially as these technologies proliferate internationally through NATO and other allied nations.

The analyst emphasizes that while AI has legitimate productivity applications, the aggregate trajectory warrants serious concern. Practical defensive measures include removing personal data from data broker networks and advocating for legislative protections against government surveillance. The fundamental question posed is whether this technology infrastructure ultimately serves users or governments and corporate interests—a distinction that may define geopolitical power dynamics in the coming decades.

Key moments

3:00Pentagon Demands Rejected by Anthropic

We cannot in good conscience acquiesce. They wanted Claude to analyze data on Americans. That data included geolocation, web browsing data, and personal finance information purchased from data brokers.

7:00Maven-Claw System Capabilities in Iran Strike

Striking 1,000 targets in 24 hours in Iran was only possible with a dedicated AI system called the Maven Smart system. The system allowed one artillery unit to do the work of 2,000 staff with just a team of 20 people.

12:00OpenAI's Opportunistic Military Deal

In a baffling move a couple of hours after the fallout between Anthropic and the US government, Sam Altman said that Open AI would take the deal that Anthropic rejected. He himself called the deal 'sloppy and rushed.'

15:00Mass Public Boycott Movement

Over 2.5 million people have taken action, either by cancelling subscriptions, sharing boycott messages on social media, or signing up via quitgpt.org. A recent Quit GPT Instagram post has more than 36 million views.

20:00Pentagon's Surveillance Risk Assessment

What if this software went down, some guard rail kicked up, some refusal happened for the next fight and we left our people at risk. We're potentially so dependent on a software provider without another alternative that has the right or ability to shut it off.

Get AI-powered video digests

Follow your favorite creators and get concise summaries delivered to your dashboard. Save hours every week.

Start for free