Fermi letter to Dirac.

Source: "From Florence to Fermions: a historical reconstruction of the origins of Fermi’s statistics one hundred years later", by Roberto Casalbuoni arxiv.org/pdf/2602.04484…

Atoms >>> Bits

Everyone who works in “AI or Techbio” should be forced to shadow in a lab for at least 1-4 weeks to see what the next steps after you have a structure on your computer.

And that’s just boring brute force research, not exploratory research with new hypothesis.

Rearranged the front-facing physics/math stack.

In January 2025 my mom’s lung cancer spread to her brain. The drug with the best chance of helping her was approved in China but not here.

A doctor told me, almost as an aside, that if she were Chinese she could buy it at a pharmacy.

So I flew to Shanghai and came home with 26 boxes in my luggage. That was 19 months ago. My mom is still here, still working, and in her words has had an amazing three years.

The drug was approved in China in 2023. The FDA approved it in July 2025. It still is not for sale here. I know of at least 30 Americans who have made the same trip.

I have been scared to tell this story because I worried it could cost my mom access to the medicine keeping her alive. But the problem matters more than my silence. These drugs need to be developed and available here, and that means FDA reform.

Art Allen at KFF Health News told our story this week. It also ran in The Washington Post.
kffhealthnews.org/health-industr…

Good to see this since atoms >> bits.

This review of Federer's "Geometric Measure Theory" by someone who was struggling to understand it for years is pure gold.

"It was only a couple of years ago, after writing a 1500 page book (now 2015 pages) on the foundations of differential geometry that I finally managed to get the gist of the first 8 pages."

amazon.com/portal/custome…

There were two hilarious Amazon review of Jackson's Electrodynamics from 2003 that I remembered. The magic of AI and the wayback machine helped me retrieve it.

"Jackson: Can't Live With It, Can't Live Without It!!!, June 28, 2003
Reviewer: Gregory Bravo (Buffalo, NY United States)

The title of my review just about sums my opinion on this "classic" grad electrodynamics text. The book kind of [stinks] as a textbook, but there is nothing even remotely close to it in scope out there.

So like a previous reviewer said: "Jackson's here to stay; GET USED TO IT!!

(1) The problems are hard. Damn hard. Someone else already said that, and I agree. What I WILL add, however, is that some of the problems are also simply STUPID and a waste of time, offering or enhancing physical understanding very little if at all. (Don't get me wrong-- there are some problems which, while hard, are also pretty darn cool. Unfortunately, there are too many of the other kind, too.) The type of problems I am talking about are of the following ilk: "Prove the following six-term vector identity;" "Re-derive equation #72 for a transverse magnetic field'" "Prove equation #27." Quite simply: WHO CARES!?!

(2) While the volume is pretty encyclopedic, it is often hard to follow. Jackson often simply states things in the text without explaining where they come from, how they are derived, or why they are important,--- for example, as I read the text, I began to hate the two words "we see," which are used is cases like (paraphrasing now) "Therefore, we see the following relationship holds"---when it was not at all clear to me where the heck this relationship was coming from! I often felt stupid because, in fact, I often did NOT "see" at all!!! In those cases, I began to just say to myself "well, if Jackson says so," and then moved on. Whether I am stupid or not is open for debate, surely, but nevertheless I would have liked a little more pedagogical assistance--- which, in my mind, does not seem too unreasonable a thing to expect in a *gasp!* TEXTBOOK, of all things!! (We aren't taking about a paper in a peer-reviewed journal now, are we?? We are talking about a textbook for people who don't know, but are trying to learn, E&M.;)

(3) This is a comment on the Third Edition (ie. "the Blue One") versus the Second Edition (ie. "the Red One"). In the Red One, Jackson uses Gaussian units throughout, which seemed to work well for me when I first started using the book on my own several years ago. Then, when I recently took a grad course on E&M; the professor used the Blue One-- wherein Jackson switches between Gaussian and SI units, depending on which chapter you are in. He explains his rationale-- which makes sense on the surface (you really need to know both units as a working physicist)-- but in practice all this did was make learning 20 times more difficult! First you get used to the way the Gaussian system works, then suddenly all the constants and definitions change and you are in SI units in the third chapter, then you are back in Gaussian units, having to remember that certain things and definitions change and certain stay the same, all the while trying to incorporate new concepts into this body of knowledge, then you're back in SI and so on. It is simply the biggest blunder that he could have made in expanding to the new edition. While there are new problems and so on in the Blue One, when I actually wanted to learn something I gave up on the Blue One and pulled out my handy Red One.

And finally...

(4) A (very minor) point: The actual SIZE and FEEL of the Red book itself is much better than the newer Blue One. The Red One feels sleek and self-contained and fits well in your hand and your backpack. The Blue One, on the other hand, is thicker, wider, bulkier and more unwieldy, and borders on the "Is there some way I can leave the book in school rather than dragging it home?" category. In short, the Blue One feels like a burden, while the Red One feels sleek and cool. Using the Red One makes you feel like you're doing something cool, slick and important, while using the Blue One makes you feel like you are burdened with the weight of physics (I'm sure all of us have felt both at some time or another-- but if you're gonna publish a textbook, go with the first!)

Well, I guess that's about it for now. I hope it was helpful."

Another hilarious one: "Yes, it is very mathematically demanding. Yes, some of the discussions (particularly towards the ends of chapters) are thoroughly inpenetrable. And yes, each chapter features a few problems of the type "Show that (horrendous expression)=(even worse expression)=(multi-line, triple sum of modified Bessel functions expression)." But with a serious effort you'll make it through the first 3/4 of every chapter and >half the problems; the remaining parts are usually specialized topics anyway. The hardest part of studying this text is simply the large amount of time you need to invest; it doesn't read like Griffiths' book. And what did you expect, E&M to be easy?"

Correct headline: Bank agents are detecting customer agents and the two are disputing transactions.

China has gotten ahead of the US in biotechnology, largely by making their clinical trial system more effective. I covered this in a @nytimes article.

Now, American patients are travelling there to have access to drugs they can't access at home. It's high time the US woke up and implemented much needed reforms.

China has selected its first group of young researchers to receive funding for up to 15 years, in one of the longest research grants on offer to scientists globally

go.nature.com/4z4Q3pw

I find books with blue covers and gilt lettering particularly beautiful.

Perhaps the most moving part of John Nash’s story is how his friends, family and colleagues came together to fund his mental health treatment because they wanted to save his beautiful mind.

Selling AI bio-risk without discussing countermeasures is like selling Reagan's Star Wars without discussing countermeasures.

I think there's a lot of misunderstanding about AI causing dangerous runaway biorisk. Here's what I said to a worried colleague:

I think there are three aspects of this that are important. Firstly, there's lots of things in biology that are just unknown and for which we don't have the data, and the attrition rate is so high that neither humans nor AI can do very well. Drug discovery is a spectacular example - most things just don't work, and we are always swimming in a sea of ignorance.

Secondly, lab work is extremely messy and prone to false positives and negatives and things just not working. AI can help accelerate some of this work - I have been part of such projects myself - but we are very far from AI being able to autonomously orchestrate any kind of risky work in a way that poses unknown dangers, except those that would arise from accidents (glassware and robotic arms breaking, chemicals being mixed unintentionally).

Thirdly, all the autonomous biological discovery that I am seeing is almost exclusively computational and falls into two categories: discovery with a human strongly in the loop, and discoveries where AI has sped up an already well-understood process.

That being said, I agree that it's a good idea to sandbox or gate AI for biology or chemistry experiments with a strong human-in-the-loop component, but more so in my mind because it could spend valuable capital chasing minor goals or cause dumb accidents rather than design a supervirus autonomously or anything like that.

I am also in favor of pausing new model development for, say, a year while we discuss future potential biorisk and agree on international safeguards, the way it was done at Asilomar with recombinant DNA technology in the 70s.

I think the bigger issue in some sense for me from an optics standpoint is that if tomorrow a bad actor designs a gain-of-function mutation "using AI", even if it just means speeding up an existing process or doing literature searches faster or better - basically stuff they could have done without AI - the alarmists are going to yell, "We told you so!"

Just watched the @netflix Instadocs about the @OpenAI cyberattack on @huggingface. Super disappointing, as they went for maximum anthropomorphism of the technology and full doomer mode.

I spent more than two hours with them trying to correct all their mistakes and ground this in t agents are technological systems, not aliens. But it looks like the drive to scare people for maximum views was too strong.

Also, they didn’t mention the lack of containment and monitoring of the agents, how closed-model safeguards blocked our defensive work, or how we defended ourselves with AI using an open model. And that we still lack a lot of information and transparency, including full agent traces and that companies need to remain accountable for what their agents do.

It ends up being a fun sci-fi movie though!

I *wish* the bottleneck in drug discovery was new ideas.