americanaffairsjournal.org/2… Nuclear energy is not just one option among many within a diversified clean energy portfolio; it is strategic sovereign infrastructure, analogous to missile defense, civil aviation control, or central banking. It is a system whose development cannot be delegated, whose coordination cannot be crowdsourced, and whose permanence transcends commercial return.

Did volcanic eruptions ruin China’s Ming dynasty and undermine the Qing? sc.mp/y3lnd?utm_source=twitt… via @scmpnews

profgmedia.com/p/the-weight The AI narrative has turned negative. Energy costs are up, jobs are (supposedly) going away...if the technology advances as promised, it presents a threat to humanity. The GLP-1 story offers something Silicon Valley used to produce in abundance: a hopeful vision of the future.

Why can’t we build infrastructure in Britain? A look at the real constraints—from high costs to short-term politics—and why lasting solutions require a whole‑economy shift. bit.ly/43BDXWE #Infrastructure #UKeconomy #EnergyPolicy #sustainableeconomy

heritage.org/markets-and-fin… and source: papers.ssrn.com/sol3/papers.…

AI Can Do Math, but Is It Really Math? by Daniel Kipnis wsj.com/opinion/ai-can-do-ma… via @WSJopinion If nobody understands a mathematical proof, is it a proof at all?

"What did you do by age 15?" In 1930, the most common answers were “Read a book for fun” and “Learned how to cook.” In 2026, the most common answers are “Watched porn online” and “Went on a diet.” tinyurl.com/4f88vtwh

Gory details of KPMG’s ‘Redefining Excellence’ paper, which is atrocious AI slop. 40 of 45 citations can’t be corroborated. It’s a historic own goal. gptzero.me/news/investigatio…

This is wild—and likely a sign of things to come as we transition to a web that is optimized for bots more than humans. theatlantic.com/technology/2…

The cost has more than doubled in real terms, and you are trying to pretend we have underinvested. Utterly shameless. Truly disgusting.

Clinical psychologists increasingly emphasize oppression and victimhood over personal agency - a shift that could undermine one of psychology’s most effective treatments: cognitive-behavioral therapy (CBT). stevestewartwilliams.com/p/t…

One way to think about the current structure of govt is to see it as an information processing and decision making system. As such, it relies very heavily on serial processing. That may work OK when the information flows are light (as e.g. in the 18th century), but it just gets overwhelmed when the flows become heavy (as they are today). Advised response: more 'lateralisation' (a more brain like structure) and more parallel processing capacity please!

Thames Water is not just a corporate failure—it’s a failure of politics and regulation. Delay has raised costs, weakened accountability, and made real reform harder. Special Administration remains the right answer. Read my latest paper here: bit.ly/4urnCyJ

Together with UC Berkeley we are announcing the laser phase plate - a breakthrough in atomic resolution imaging. This is the brightest continuous wave laser in the world, 100 million times the intensity of the surface of the sun. Phase contrast plays an important role in microscopy, but it was thought close to impossible for electron microscopy, where it would require interfering with an electron beam. Holger Mueller and Robert Glaeser proposed exactly this using a standing wave laser. It has taken over 15 years to make this a reality. Biohub partnered with UC Berkeley and Mueller to support this work and to engineer and build the technology. Contrast has been the critical barrier to achieving atomic resolution imaging of the cell. In cryo-electron tomography, a cellular imaging technology that uses electron microscopy, the low contrast makes it impossible to resolve anything but the largest proteins within their cellular context. The laser phase plate removes that barrier. With advances in AI this breakthrough in contrast will start to open up a new frontier in structural biology, that will allow us to see the molecular machines of the cell, and how they assemble into far more complex and dynamic systems, and understand how they work.

bostonreview.net/articles/kn… Knowledge Collapse. AI companies are racing to mechanize mathematics. Where does that leave human understanding?

technologyreview.com/2026/06… Why China is betting on big nuclear reactors. While companies in the US chase smaller footprints, there are a lot of new large reactors going up in China.

New preprint! We introduce a new benchmark, SciConBench, with 9.11k scientific questions derived from Cochrane Systematic Reviews. We find evidence that frontier AI agents **cannot** synthesize scientific conclusions well. A thread 🧵 w/ @hayounggjung, @korolova & others

I was 26 years old when Peter Lynch handed me this. April 28, 1983. I was the auto and retail analyst at Fidelity. Peter was in his prime, on his way to building the greatest mutual fund track record in history: 29.2% annual returns for 13 YEARS STRAIGHT, growing Magellan from $18 million to $14 billion. The Babe Ruth of investing. I'm looking at the principles he had typed up on a single sheet of paper that I've kept in my files for 42 years and I believe now is the perfect time to revisit them again. Let me walk you through a few: Rule 1B: "You need an edge to make money. Do not rely on a combination of hope and good luck." Today's retail investor has no edge. He has Reddit, Robinhood, zero-DTE options and a TikTok algorithm pushing him into whatever stock just ripped 200% the day before. That's hope and good luck wearing a fancy costume. Rule 1E: "Purchase stocks like one would purchase a business." Tesla trades at over 360 times earnings on a business deteriorating in real time, Oracle has $206 billion in liabilities against $39 billion in equity, MicroStrategy is a leveraged Bitcoin holding company priced like a software firm, and don't even get me started on SpaceX, that piece of garbage you'll be able to trade tomorrow... Nobody in their right mind would buy these as actual businesses. They buy them as stories, narratives, and lottery tickets. Peter would have called it the same way I do - these are not investments. They are speculations. GAMBLING. Rule 1G: "Study the balance sheet and cash flow statement." The hyperscalers spent over $380 billion on AI capex in 2025. Goldman says the measurable productivity payoff does not arrive until 2027 at the earliest. Oracle just reported NEGATIVE $23.7 billion in free cash flow for fiscal 2026 while borrowing at a pace that would make a leveraged buyout firm nervous. The cash flow statements are screaming but nobody is reading them. Rule 1I: "Avoid the long shot." This one cuts the deepest. The entire market has become a long shot. OpenAI is projected to post roughly $74 billion in operating losses in 2028 ALONE while priced for transformation tomorrow. Bitcoin treasury companies are multiplying off thin air. The retail investor of 2026 is making one long-shot bet after another and calling it a portfolio. Rule 3A: "When the fundamentals change, sell your mistakes." Tesla's fundamentals have changed. California registrations are down 24% year over year and inventory days went from 10 to 27. Musk himself admitted on the last earnings call that Hardware 3 cannot achieve unsupervised FSD, breaking a promise made to 4 million customers. The fundamentals have screamed change. But the stock is still at $385. The mistakes are not being sold. They are literally being doubled down on with leverage. Rule 3I: "A 30-50% profit in 12 months is great. Mediocre in three years." Today's retail crowd expects 30-50% in a WEEK. Then they wonder why they get wiped out the second the hype stops. And my favorite - Rule 3J: "Develop your own style and stick to it." That is the entire game right there. I developed mine sitting across the hall from Peter Lynch in 1983, watching him work, reading his notes, getting my own research handed back to me covered in his pencil marks. Then in 1984, my first full year managing money, I ran the #1 mutual fund in America. The Fidelity Overseas Fund was top 2 for the next six years running. I did not get there by chasing narratives. I got there by following the sheet of paper you are looking at right now. 42 years later, this single page contains more wisdom than every Fintwit thread, CNBC segment, and Wall Street price target combined. Peter retired in 1990 with the greatest mutual fund record in history. Then he sat down and wrote books explaining exactly how he did it. Only a few "investors" these days read them. And almost nobody is reading the balance sheets, the cash flow statements, or studying actual businesses today either. They are chasing AI, crypto, and whatever pumped yesterday. The wisdom on this page is timeless and it's more important than ever.