John 🧂🔋

We’ve been recognized as a Tier1 energy storage supplier by @BloombergNEF for the first time ever and are 1 of 5 US companies & 1 of 2 alternative battery technologies on the list! This list is curated based on rigorous standards each quarter – we couldn't be prouder of our team!

In 1991, Patrick Stewart surprised Star Trek creator Gene Roddenberry with a special birthday performance aboard the USS Enterprise-D. Dressed as Captain Jean-Luc Picard, Stewart sang his way through the alphabet, delivering playful lines of praise while dancing across the iconic Star Trek: The Next Generation bridge. Originally filmed as a private tribute during the show’s fifth season, the footage wasn’t intended for public viewing. Years later, it resurfaced as a bonus feature on the Season 5 DVD, giving fans a rare look at Stewart’s playful personality and the cast’s appreciation for the man who created Star Trek.

31 October 1973. 1st episode of the 26-part classic ITV documentary series World at War was broadcast. It was narrated by Laurence Olivier with music by Carl Davies. It cost £900,000 (worth £11 million today).

Trump family friends doing business with Russia, which it turns out secretly controls the $25 billion Alaska LNG project. ft.com/content/e9cfbfbb-e65a…

Thought I’d give his book a go…

"Guy answers an AI job interviewer in pure gibberish and she has a full meltdown"

There is a fundamental mistake creeping into the AI debate: We are confusing accountability with collective decision-making. If an AI company builds a powerful system, that company should own the consequences of building and shipping it. Not “society.” Not an international committee. Not a crowd of stakeholders who cannot see the model weights, understand the architecture, reproduce the evals, or observe what is happening inside the lab. The company! Start from first principles. Responsibility should sit as close as possible to: the information, the decision, the upside, and the downside. The people building frontier AI possess vastly more information about their systems than the general public. They decide what gets trained. They decide what gets deployed. They capture the economic upside. Therefore they should bear responsibility for managing the risk. This is not a radical governance model. It is how civilization has handled technological progress for centuries. Boeing engineers do not hold a national referendum before changing an aircraft. A pharmaceutical company does not ask the population to design its clinical trial. A nuclear plant operator does not crowdsource reactor procedure. Companies build expertise, establish safety margins, test their products, slow down when necessary, and accept liability when they fail. AI should be no different. If OpenAI believes autonomous recursive self-improvement is currently too dangerous, OpenAI should not deploy it. If Anthropic believes a capability threshold requires six additional months of alignment work, Anthropic can wait six months. If Google believes its next model is unsafe, Google should not ship it. You get the point? That is what taking responsibility means. But transferring the pacing decision from builders to “society” creates a completely different problem: the information problem. The further decision-making moves from the people who actually understand the technology, the less information the decision-maker possesses. And when knowledge is asymmetric, collective decision-making does not magically aggregate wisdom. It can aggregate ignorance. The wisdom of crowds works when individuals possess dispersed, relevant information and their errors are reasonably independent. Frontier AI has almost the opposite structure. The relevant information is extraordinarily concentrated. A tiny number of researchers understand the systems at the frontier. The broader public necessarily operates through abstractions, headlines, political narratives, fear, lobbying and second-hand interpretation. Adding more voters to that decision does not necessarily add more information. Sometimes it simply adds more noise. There is also an incentive problem. If responsibility is collectivized, accountability becomes diluted. A company can eventually say: “We followed the standard.” “The regulators approved the framework.” “The international body signed off.” “The process was democratic.” Everyone participated. Nobody was responsible. That is backwards. We should want the CEO signing the release to know: If this goes wrong, it is our decision. That concentrates attention wonderfully. Of course we need laws against externalities, common measurement standards, incident reporting, national-security protections and liability where appropriate. But those are guardrails. They should not become a steering wheel. Because technological progress is fundamentally a process of decentralized experimentation. Thousands of decisions. Thousands of failures. Thousands of iterations. You cannot centrally optimize a search process whose destination is unknown. The danger is that in trying to eliminate every theoretical downside before moving forward, we eliminate enormous amounts of unknowable upside as well: - cancer treatments not discovered, - energy breakthroughs delayed, - scientific discoveries postponed, - productivity gains never realized, - defensive technologies developed too slowly. There is no risk-free path! Slowing technological progress has risks too! And those risks are almost never represented by the people demanding the slowdown because the victims of foregone progress are invisible. Nobody knows the name of the patient whose cure arrived ten years late. So the governing principle should be remarkably simple: Those who build should be responsible for what they build. Let companies pace themselves when they encounter genuine danger. Make them disclose material failures. Make them bear consequences for negligence. Create objective technical standards where coordination genuinely helps. But resist turning technological progress into a global permission structure. Because once everyone gets a veto over progress, progress eventually gets a veto. The frontier should have guardrails. It should not have a committee holding the steering wheel. @JensenHuang @elonmusk @DavidSacks

Russia has to be stopped by the West at all cost. Culturally we are not the same and they want to see our systems fail. Culturally and philosophically they are much closer to Muslims than us westerners. Both are barbarians that are ruled (and want to be ruled by) strength and fear, rather than virtues.

$EOSE | Frontier Power USA (FPUSA) Frontier Power USA officially announced its leadership team today. I think the resumes are worth paying attention to. FPUSA says the team has collectively developed 20+ GW of energy storage across ERCOT, CAISO, PJM & NYISO. Jay Bellows - CEO Former CEO of KORE Power & President/CEO of Northern Reliability. At KORE, Bellows led one of the early vertically integrated U.S. storage businesses spanning domestic battery manufacturing, engineering, integration & long-term service. Northern Reliability deployed 1,000+ energy storage systems globally across utility, government, developer & C&I applications. Bellows also has history around DoD / U.S. Navy energy resilience, including Northern Reliability’s selection for an EPRI microgrid project at Naval Base Ventura County funded in part by the DoD ESTCP + California Energy Commission. Mark Klein - Head of Development Co-founder & first employee of Broad Reach Power. Broad Reach became one of the largest battery platforms in Texas before its 2023 sale to ENGIE, with much of ENGIE’s current North American storage platform tracing back to the business Klein helped build. Klein also spent 12 years at PacifiCorp Energy / Scottish Power in senior commercial development & trading roles. Nitin Gupta - Head of Capital Formation Former SVP of Finance & M&A at Broad Reach Power. Previously CFO of Voltwise Power, backed by Sandbrook Capital, and earlier VP at GE Energy Financial Services. His experience spans raising and deploying project + portfolio capital for utility-scale storage across the U.S., UK & Germany. Jay Legault - VP of Operations 13+ years in energy with an interesting background coming directly from the nuclear sector. Previously Civil Maintenance & Radiation Protection Coordinator at Ontario Power Generation’s Darlington Nuclear Plant. His background is centered around construction, operations, execution & regulatory compliance …bringing nuclear-grade operating discipline into FPUSA’s storage buildout. Lineup summary: Bellows → technology / manufacturing / operations Klein → origination / development Gupta → capital formation / financing / M&A Legault → construction / operations / regulatory execution This isn’t simply a sales organization. FPUSA is assembling people who have already developed, financed, constructed, operated and transacted storage at scale. And they’re reconnecting pieces of the team that helped build Broad Reach. Klein & Gupta previously worked extensively with the principals of Stella Energy Solutions, now FPUSA’s designated execution partner. That group delivered its first 17 utility-scale battery projects (~350 MW) in under 36 months at Broad Reach. The quote today says it best: “We have the technology, the manufacturing, and the capital in place, and we now have the team to convert that into gigawatt-hours on the grid.” The next thing that matters is conversion. Frontier Power USA announcement: frontierpowerusa.com/frontie…

Thanks 🙏, @Strive. Very happy to have $ASST as the largest part of my portfolio.

Zinc Hits Four-Year High As "Extremely Thin" Physical Supply Fuels Squeeze zerohedge.com/commodities/zi…

wwow.. AI just turned red alert 2 into 22mins epic film