In this episode, Sunny Ray and his co-host open by celebrating securing the Allin.com domain before diving into a wide-ranging AI news roundup. They discuss the US government's order for Anthropic to suspend worldwide access to its Claude model over national security concerns, debating whether a jailbreak vulnerability or a broader surveillance dispute was the real cause. The conversation shifts through UFC 250 fight commentary, Bernie Sanders proposal to have the government take fifty percent equity stakes in major AI companies, and Coinbase and Visa launching payment rails for autonomous AI agents. They cover Tesla Optimus V3 humanoid production, SpaceX acquiring Cursor for sixty billion dollars, Snap's new AR glasses, and a leaked OpenAI financial report showing large losses despite billions in revenue. The co-host closes by demoing a personal AI system he built, called Sovereign, consisting of Pi, a personal AI infrastructure that knows his life data, and Hermes, an execution layer built on Claude Code and Codex that can browse, scrape, and build on his behalf.
Sunny Ray and his co-host unpack Claude's government shutdown, AI agent payments, and a homemade personal AI second brain.
What actually happened with Claude getting banned?
The US government ordered Anthropic to shut off worldwide access to Claude, citing national security. I think it is total nonsense in my opinion. Some reporting suggests a jailbreak was found, possibly by someone at Amazon, but every model gets jailbroken within an hour of release, so it feels like an unsolved industry problem being used as a pretext.
Do you think the jailbreak was really why they shut it down?
I do not fully know what to think. The most cited read is that the jailbreak was the pretext, refusal to build surveillance weapons was the actual context, and the export control directive was just the mechanism used to enforce it.
What is your take on Anthropic's Dario Amodei compared to other AI CEOs?
Unlike Sam Altman or Elon, I actually find myself paying attention when Dario talks. When he stood up to Trump, he was clear that he would not let the government spy on Americans or let AI systems make kill decisions without a human in the loop. I found that honorable, even if opinions on him are split right now.
What do you think about Bernie Sanders wanting the government to take fifty percent of AI company revenue?
It does not surprise me, he is a socialist, so that is exactly what he wants. But I think Trump would actually take one hundred percent by trying to personally own these companies or place himself on the board. That kind of move is very much in his character.
Why do you think SpaceX acquired Cursor for sixty billion dollars?
Elon now has models that are quite strong, and Cursor sees how millions of developers actually use ChatGPT, Claude, and Gemini in the background. That makes it the perfect acquisition for him to learn who is winning and build something better. He is playing a different game, swimming in from every direction with satellites, data centers, and now coding tools.
Can you walk us through the personal AI system you built?
It is called Sovereign, made of two parts. Pi, personal AI infrastructure, is a second brain that knows everything about me down to my heartbeat, from every email to every WhatsApp message, plus my goals and net worth. Hermes is the execution layer built on Claude Code and Codex that scrapes websites, builds dashboards, and does anything I can do on a computer just by me speaking to it.
Building something daring? Sunny talks to founders like this every day. Fifteen minutes to see if your story belongs on the stage.
Claim your pre-interview