killjoywashere

Denmark is in a chronic baby shortage [1] and people in Western democracies are having less sex generally [2]. So, yay, less HPV. Go get vaccinated [3]. Unfortunately, there are some pretty significant (and sad, yes, sad) confounders.

[1] https://www.sdu.dk/en/nyheder/faldende-fertilitet

[2] https://www.google.com/search?q=western+democracies+decreasi...

[3] https://pmc.ncbi.nlm.nih.gov/articles/PMC10399474/

I have never worked for a company. I have always worked for the government, the taxpayers, quixotically according to a lot of people on the internet, even, dare I say, the citizenry. Most people think these are cush jobs.

I spent decades worked way more than 996, on ships, ashore, in medical school, in residency, on clinical staff while doing entirely uncompensated research. Now I'm a subspecialist physician living in the Valley. I have never worked this little and enjoyed such a high standard of living. One of my seniors said "You don't have to work 2.5 jobs anymore. Just work 1.25 jobs". I work with teams across the spectrum of businesses to figure out how to build the business lines and I see the challenges small companies have. I really do. Not least of which is how the big companies have stacked the deck against new entrants.

Now that I do have some free time I spend it helping my wife build her business, I'm essentially her cofounder. Been incorporated for 8 years now. We think about motivating employees, paying them fairly, the breath-taking amount of money consumed by SaaS, rent, health insurance, travel costs and how that makes it hard to pay employees more. We think about motivating customers and charging them fairly. We see the mind-reeling amounts the big companies charge and then give customer discounts that effectively curb the competition. I see how they get their employees to work harder.

There are two fundamental rules in business:

1) If you're not making money, you're losing money.

2) Don't run out of money.

We watch the end-of-month profit margin going up and down like a rollercoaster. Some months, yeah, "This is great". Some months "Oh, oh, we cannot keep doing this".

We had one employee who really took this whole "I don't have to work ... hard" to heart. She would charge an hour for filling out her timesheet. She consumed her annual sick leave and accumulated PTO in her first 6 weeks. She would bail on scheduled work. Customers loved her but she was literally a net cost to the company money. How? Fixed costs. Overhead is real. Had to let her go. Honestly wasn't a hard conversation with her (she actually never returned some equipment, flat out stole from the company). What was hard was figuring out how to cover those customers and explaining to them why their favorite face of the company was gone.

You want to live a happy, ethical life? Live within your means. But that also entails having the means needed. And everybody else gets a vote. If you live in the US: the whole world wants your quality of life. Even if it's just 10% of the rest of the world, that's still double the entire US population, who are working 996.

Chatroulette was on to something. But is it still full of men exposing themselves?

If you've ever used liquid nitrogen to snap-freeze tissue in gluteraldeyde for electron microscopy, the problem is readily apparent: you can't get the heat out of large chunks of meat fast enough. And by large, I mean 1 cm cubed. 0.5 cm cubed, maybe.

A buddy from Bridgepoint made a point to me once: maybe we're not number one in every industry, but find me a sector where we're not in the top 10.

This. As someone who spends a lot of time thinking about national security problems, it's hard to convey how badly the 1980s MBA education failed us as a country. "Greed is good." Sure, Gordon. For who? Who's greed is good for who? How about the people of the country labor for the long term wealth of the nation? How about we all work for the long term wealth of the planet?

In case anyone thinks assistants serving others can't have some incredibly dystopian consequences, The Star Chamber podcast has an incredible 2-part series, * With Friends Like These ...* describing a case that boggles the mind.

Part 1: https://www.youtube.com/watch?v=VVb7__ZlHI0 (key timestamps: 31:45 and 34:3)

Part 2: https://www.youtube.com/watch?v=vZvQGI5dstM (key timestamp: 22:05)

If you're like "Woah, this seems kinda disconnected, I'm missing context..." Uh, yeah, there's so much context.

Here's the link that most critical bit in Part 2: https://youtu.be/vZvQGI5dstM?feature=shared&t=1325

And if you listen to the whole thing, here's the almost innocuous WSJ article:

Here's the WSJ article that put it into the press: https://www.wsj.com/politics/national-security/workplace-har...

I happen to have a stereo microscope at my desk, so I put my Pixel 9 under there. At 100x mag (10x ocular x 10x objective) it looks like there are 3 layers: as I move my head around slightly (so the image is moving over my retinas), the blue moves faster and the red almost stays still, with green somewhere in the middle.

David Hilbert was one of the greatest mathematicians of all time. Many of the leaders of the Manhattan Project learned the mathematics of physics from him. But he was famous long before then. In 1900 he gave an invited lecture where he listed several outstanding problems in mathematics the solution of any one of which would change not only the career of the person who solved the problem, but possibly life on Earth. Many have stood like mountains in the distance, rising above the clouds, for generations. The sixth problem was an axiomatic derivation of the laws of physics. While the standard model of physics describes the quantum realm and gravity, in theory, the messy soup one step up, fluid dynamics, is far from a solved problem. High resolution simulations of fluid dynamics consume vast amounts of supercomputer time and are critical for problems ranging from turbulence, to weather, nuclear explosions, and the origins of the universe.

This team seems a bit like Shelby and Miles trying to build a Ford that would win the 24 hours of LeMans. The race isn’t over, but Ken Miles has beat his own lap record in the same race, twice. Might want to tune in for the rest.

I look forward to the first LLM-designed massively multifamily dwellings. I'm sure no one will fall through any floors.

Bob?

IdenTrust participates in the US Federal PKI ecosystem, so they likely have strong incentives to charge exorbitantly. Those free certs are probably meant to facilitate development of gov-specific capabilities by random subcontractors long enough to figure out how to structure a contract mod that passes the anticipated cost onto the government.

Don’t hate the player, hate the game.

Fun fact: these laser pulses are so short they are no longer a single wavelength. They have a spectrum due to the uncertainty principle. And at this short of a time scale, it’s pretty broad.

I've met Steve on a number of occasions, smart guy. I also work in this space and have definitely read a lot of the same books he has, along with many of his excellent articles. Alas, I'm pretty sure he would never read my articles, unless he got cancer, which I hope he doesn't.

I do want to pick up where he left off: the interface between the US and China, and specifically look at how China has invested. I've spent some effort on this forum making the point that our system has left some critical vulnerabilities that the Chinese have leveraged, e.g. (1,2).

It's worth understanding that Xi Jinping has been working hard on this problem set, along with his predecessors and many around him for a long time. To really understand his whole-of-economy approach, I highly recommend Hank Paulson's Dealing with China (3). He has and maintains a narrative of literally going from a boy in a cave to the leader of the largest nation on Earth. Much like the narrative arc Churchill maintained for himself (the Prof Blank mentions), Xi would see science as a component of the tapestry, but not the whole story.

Xi is also using the Belt and Road Initiative for massive effect, see the maps in (4). The US has started to pay attention with renewed investments in the region, e.g. (5) but Xi has a decade head start and a political base that could be characterized as relatively stable compared to the current US administration.

As my time is limited, I'm appending a reading list at the end for those interested (6 to end). Suffice to say, yes this is how we became a science superpower. But it ignores how our parochial incentives and belief in American exceptionalism morphed in the American narcissism (14) this is very likely to doom the American experiment without significant effort on the part of the American population to come together. Unfortunately, I fear the fracturing of the population is too far gone to remediate without major conflict, but major conflict in the present setting is likely far more serious than we could survive as a nation.

As a final thought, the major conflict is obviously nuclear war. We will not survive that as a nation. Thus the Prisoner's Dilemma. We are all prisoners on Earth. Even Musk's species-level escape is far from escape. The physics of deep space travel or even intra-solar-system travel just don't work out in our favor. So, how do you survive the Prisoner's Dilemma? The math answer is "there are a lot of complicated answers" ref (15) but mainly, all parties need to work toward, and signal reliably that they are working toward, stable equilibrium. Being an unreliable partner must be met with brutality, even at the cost of everyone.

(1) https://news.ycombinator.com/item?id=43655390

(2) https://news.ycombinator.com/item?id=20321493

(3) Hank Paulson, Dealing with China. We, and specifically, Goldman Sachs, and specifically Hank Paulson, taught Xi how to win. https://www.amazon.com/Dealing-China-Insider-Economic-Superp...

(4) https://merics.org/en/tracker/how-bri-shaping-global-trade-a...

(5) https://asiatimes.com/2025/02/us-revives-wwii-era-pacific-ai...

(6) Manchester. The Last Lion, the 3 volume definitive biography of Churchill, which puts the Prof's work in the largest possible context. https://en.wikipedia.org/wiki/The_Last_Lion:_Winston_Spencer...

(7) Jamie Holmes. 12 Seconds of Silence, the definitive story of the proximity fuse, a significant portio of Merle Tuve's unique contributions to the war, and the story of the founding of Johns Hopkins' Applied Physics Laboratory. https://www.amazon.com/Seconds-Silence-Inventors-Tinkerers-S...

(8) Richard Rhodes, The Making of the Atomic Bomb. Note Merle Tuve also plays a critical role in this narrative, not bad for one of those 'second rate' government labs. https://www.amazon.com/Making-Atomic-Bomb-Richard-Rhodes/dp/...

(9) Rocco Casagrande and the work of Gryphon Scientific, alas (but probably net good) acquired by Deloitte. Wayback has some of their reports: https://web.archive.org/web/20240228103801/https://www.gryph...

(10) Senior Colonel Ji-Wei Guo, and his theory of Merciful Conquest, audaciously published in the US's own Military Medicine journal https://pubmed.ncbi.nlm.nih.gov/19813351/ see also https://www.cnas.org/publications/commentary/weaponizing-bio...

(11) Bruce Bueno de Mesquita, The Dictator's Handbook. This Berkeley professor uses innumerable real world examples to illustrate how dictators effectively control their populations https://www.amazon.com/Dictators-Handbook-Behavior-Almost-Po...

(12) James C. Scott, Seeing like a State. UC Santa Cruz professor uses several extremely large examples the illustrate other ways governments control their resources. Spends a lot of time on the negative effects but certainly acknowledges the net upsides usually seem to outweigh the net downsides, but it would be good to learn how to avoid downsides when you can: https://www.amazon.com/Seeing-like-State-Certain-Condition/d...

(13) Walter Isaacson, Steve Jobs. Most interesting passage to me was the dinner with Obama where Jobs told Obama the manufacturing jobs are never coming back. https://www.amazon.com/Steve-Jobs-Walter-Isaacson/dp/1451648...

(14) H.R. McMaster https://www.twincities.com/2020/10/16/h-r-mcmaster-u-s-forei... also https://www.amazon.com/Battlegrounds-Fight-Defend-Free-World...

Again, you're generalizing. To say that doesn't work in medieval Europe is probably itself a generalization. But if you read the book, I don't think this would be a point you'd be arguing.

1) target universities have to accept the students

Yes! Which the US incentivizes by a) underfunding K-12 education, reducing the internal applicant pool, b) competing grants in a way that incentivizes PIs to grasp for cheap labor. Additionally, the individual states also incentivize this. Look at the UC's own statistics: since 2009 the highest chance of acceptance goes to foreign, ethnically Asian applicants. https://news.ycombinator.com/item?id=20321493

2) RTS-game-esque control of the grad students

Yes! Having been in the room when one of those grad students got a very stressful call from home (like broke down crying multiple times), it's definitely not all carrots. And the removal of carrots eventually looks like a stick.

3) Seems like the rational decision for ambitious grad students

Yes! No amplification needed.

The concept of freehold tenure is pretty central to the book. Not sure you could get any more on-point for the general reader looking for a book recommendation.

But since you ask: the peasant's rights to land were exquisitely bespoke. No tax collector could figure out how much one family owed versus another in another county. The rules in one prefecture of one county may have been completely unresolvable with the rules of a county a hundred miles north. Everything was negotiated family to family over generations, with rights in one place having no corollary whatsoever with the rights in another area, making the tax man's duty a fool's errand.

So, I don't your first statement "European peasants "renting" the land they farmed had much stronger ownership rights than Americans who "own" land do today." is really meaningful. Because no generalization can be made about the rights of a European peasant. That problem is the whole reason for the systems of freehold tenure that prevail today: making the territory "seeable" by the state.

Someone kill that muppet that jumps out of the margin halfway down. Please.

Recommend James C. Scott's "Seeing like a State" to learn more about the evolution of property valuation and rights. The systems of land rights in up to the 1500s-1800s were quite complex. The modern state imposed a uniform system of free-hold tenure which shifted the complexity to the downstream consequences.

https://www.amazon.com/Seeing-like-State-Certain-Condition/d...

Your source leans right-center, so probably good reason to suspect their reported top 1% AGI is low and their reported federal income tax estimate is high.

https://mediabiasfactcheck.com/tax-foundation/

No but their earnings are mainly in their companies, and those can hire fleets of tax attorneys and accountants to crush their tax burden.

Once the money is in stocks, it doesn't get taxed unless you draw on it, but the billionaires can use strategies like buy, borrow, die (which last I checked only really works if you're north of ~ $300M) to avoid personal taxes.

Einstein essentially gave up his professorship at the University of Berlin. How far into the future are you looking?

https://www.nytimes.com/1932/10/18/archives/einstein-would-q...

Much like outbreaks that never turn into pandemics, no one remembers the efficiency measures that prevent price increases.

A major incentive for hiring work out to the private sector is the impossibility of firing GS-series employees. Ultimately, whomever the elites are who happen to be taking a lap through government today are interesting in maintaining a responsive chattel workforce capable of reaping and sowing the crops of the day. They depend on their slave drivers, I mean senior HR staff, to keep them informed about how their current staff mix affects their ability to react to the crises of the day. If you have tons of highly trained agroconomists in GS billets, what are you going to do with them?

I do think there's an interesting conversation to have here though about workforce management, as someone who lives in adjacent worlds.

If you are long term greedy, like China, a great strategy to capture dominance of a discipline would be along the lines of how to boil a frog. Start by sending grad students to the top universities, ensuring they work for the PIs for cheap, bring as many of them back to China as you can, but tolerate a leaky return path so as not to stir up notice. Advertize their high post-training employment rate back to the universities to keep their valves open even as you start developing your universities internally, and eventually throttle down the outbound grad student pipeline. At some point after it's too late, the top universities, and their countries, look around, bemoan the lack of people in their discipline, and then just give up because by now they're old and tired.

Seems like something that has happened in chemistry, physics, and EE for sure. Once you start thinking this way, all sorts of things start making sense. Like maybe they looked at solar as a cheap, low threat point of entry for developing silicon fabrication capabilities. Software engineering, being a relatively soft skill, comes along for the ride.

Not sure about other fields, but if AI can take on a rapidly increasing set of fields, you start seeing this as how China primarily harvests not IP but workforce training from the global West, then technologies happen to fall out, then one day China has solved for their own graying work force at the same time they've solved for global economic dominance.

And a non-trivial contributor was the US governments (I blame the states too) defunding education.

We have been looking at Hamming distance vs time to signature for ambient note generation in medicine. Any other metrics? Lots of metrics in the ML papers, but a lot of them seem sus. They take a lot of work to reproduce or they are designed around some strategy like maxing out the easy true negatives (so you get desirable accuracy and F1 score), etc. as someone trying to build validation protocols I can get vendors to enable (need them to write certain data from memory to a DB table we can access) I’d welcome that discussion. Right now the MBAs running the hospital systems are doing whatever their ML buddies say without regard to patient or provider.

I think this discussion is missing a significant issue: why would a democratic government, of the people, by the people, and for the people, use tear gas? Why do British Bobbies carry batons?

The simplest answer, which is the basis of all police powers of the state: to prevent crimes against persons and crimes against property. Riotous mobs get people hurt, often killed. The businesses proximate to a protest, often small businesses, often under-insured due to cost constraints, are very likely to be severely affected by protests that turn violent. Major corporate storefronts can absorb the cost of damages, but the same damages to a mom-and-pop grocery could be the end of the business, no matter how much they sympathize with the community that is outraged. And let me reiterate what I started with: riotous mobs get people hurt, often killed.

Tear gas vs the baton is a lesser version of the observation by James B. Conant, president of Harvard University, in his autobiography: "To me the development of new and more effective gases seemed no more immoral than the manufacture of explosives and guns. . . I did not see in 1917, and do not see in 1968, why tearing a man's guts out by a high-explosive shell is to be preferred to maiming him by attacking his lungs or skin."

https://www.amazon.com/My-Several-Lives-Memoirs-Inventor/dp/...

Chemical Weapons Convention (CWC) (a follow-on to the 1925 "Geneva Convention") allows for the use of riot control agents (like tear gas) for law enforcement purposes.

https://www.opcw.org/our-work/what-chemical-weapon

Seems like a great job for a biochemist who wants to get into White Labs style brewing.

I have worked with Google teams as well, and they taught me a fair bit about how to be rigorously skeptical. It takes domain knowledge, statistical knowledge, data, time and the computational resources to challenge them. I've done it, but it took real resources.

That said, it's a useful exercise to figure out the plan of attack. My experience is the "juice" was mainly in "easy true negative" subclasses. They weren't oversampled, but the human brain wouldn't even consider most of that data. Once you ablate those subclasses from the dataset, (which takes a lot of additional labelling effort), you can start challenging their assertions. But it's hard.

And that said I also review a number of articles in that domain, and I haven't seen a group with stronger datasets overall.