Welcomed the Latvian defense and economic delegation to Starling. We showcased the Pathfinder-X: a lightweight fixed-wing drone with 2+ hours of endurance, built for the long-range missions Latvia is operating on its border with Russia.

I started Entropic. Power, compute, and autonomy under one roof. First cluster is live in Iceland. entropic.cc

Spoke to our data center operator. 80 sites worldwide. Almost every one of them is full.

When you think about resilience, think about the Cantor team @Official_Cantor and Howard Lutnick. Proud to work closely with their team. A real-life phoenix.

America didn’t lose ground in high-speed flight because of a lack of ideas. We slowed down because flight tests became too expensive, too rare, and too cautious. Meet Ramjet-X. Air-launched from the Quarterhorse program, it’s a low-cost, repeatable test vehicle built to fly and collect data at historically low costs and high cadence. This is how America wins: test faster, fail cheaper, and put hardware in the air every single week.

Congrats to the @XTEND_Xteam on listing on the NYSE. Mission-driven companies matter. Looking forward to their growth chapter as a public company.

It’s very obvious that autonomy software is going to get embedded on a lot more systems, aircraft, and UAV, that are not built in-house. This is a key example of that. Honor to be an investor in both Hermeus and Anduril.

Superintelligence must be as abundant and ubiquitous as air.

I expected B300 spot prices to be $6 ±10%. But I’m seeing $7+ with 95% utilization in my newCo deployed cluster. We will be expanding our compute operations.

In Las Vegas for the Commercial UAV Expo. Here’s Starling Systems: a foldable fixed-wing drone in a box. It does VTOL and hover with a 6-motor configuration. Endurance is 100 miles and it’s a Group 1. That’s very unique. Enjoy working with Ido and Starling.

$KTOS is down ~20% since mid-August and has been flat over the last 30 days despite 5 contract wins: - JASSM EMD, - a Javelin seeker deal, - Boeing JDAM LR engines, - $35m National Security award, - $20m+ SATCOM order for an Asian defense customer today. All in the last 30 days.

Big ship still needs a crew (and tugboats). These others don’t. Still a bunch of one-off systems, though. Coast Guard told me they need 30+ logins to run them.

$ONDS $33M Aran Defense acquisition prices at ~1.3x Aran's 2026 expected revenue. Accretive multiples like that are rare in defense-tech M&A right now.

🇮🇱🇬🇷

All of this is just attention over text. Amazing (but somewhat primitive) that human-invented language from 200,000 years ago (text, words, letters) became the reasoning engine for current AI models. The next leap in intelligence will be different and will likely not be using language as its substrate.

$WULF down ~50% from its summer highs. Price action is not accurately reflecting underlying progress, imo. Kentucky's PSC approved 482MW for TeraWulf's data campus earlier this week, and a separate regulatory gate from the 401MW Anthropic lease at Lake Mariner. Two large sites and approvals are now cleared. I'm not sure it's wise to be long AI labs that don't own and generate power. Intelligence is about the golden ratio of: $/1mtoken/MW

An unstoppable force.

Golden age of space travel

Yes. Truth to this. But AI infra has several ways to mitigate. Including under appreciated BHM power generation. Being anti-AI is the new, new thing. “occupy Wall Street”. Vibes. It will come and go. Intelligence is a universal good and it’s here to stay.

Earlier this year, my wife spontaneously discovered she had a cavernous malformation deep in her thalamus that had been hemorrhaging asymptomatically. We consulted with doctors across the country for the best treatment. Due to the extremely sensitive location of her pathology in the thalamus, the doctors recommended to do nothing and just observe because surgery risk was very high. The trade off, however, was that if not operated, this grouping of blood vessels deep in her brain could hemorrhage and lead to terminal damage. When we asked a neurosurgeon in LA what he would do, he replied: "Just hope for the best". Through a mutual friend, we learned about @BillAckman's daughter situation earlier this year and re-assessed the risk-benefit of a deep brain surgery. Although the details were vague, we understood Lucy's hemorrhage could have been caused by the the same or very similar situation as my wife's. Through various referrals, we landed with a modern day DaVinci of brain surgery Dr. Michael Lawton @mtlawton at the Barrow Brain and Spine Institute (@BarrowNeuro) in Phoenix, Arizona. Dr. Lawton and his team were the only ones with the knowledge and skillset (having completed over 1,000+ similar surgeries) to have my wife's hemorrhaging pathology removed with the best chances of no terminal impairment from surgery. On June 14, the surgery was completed successfully. Here is a published video of the surgery performed by @mtlawton: x.com/mtlawton/status/206982… Had it not been for appreciating @BillAckman's situation and the unfortunate circumstances it resulted for Lucy, we perhaps not have proceeded with surgery and, unknowingly, put my wife in similar path to a sudden hemorrhage. In Judaism, the highest act is that of pikuach nefesh, "saving a life". Indirectly, Lucy's story may have saved my wife's live. Thank you. And after observing my wife's recovery, I'm confident that Lucy will recover in full with time. The brain can work wonders. Sending you all the best. -- The @BarrowNeuro institute opened its doors 65 years ago, and was established by Dr. John Green, whose wife had a brain tumor complication that led to her death. Had it not been for Barrow Institute, delicate brain surgeries would be out of the realm of possibilities. It's hard to ignore the parallels of Bill's decision to establish the AOI today. I'm sure will look back 65 years from today and count the many lives its saved. The brain remains a mysterious frontier. Organizations, private companies and team's like these are the pioneers that continue to push brain research and technology forward. In 5-10 years, the brain will have its Apollo moment.