This is a truism that you see repeated both by hiring managers on LinkedIn, and by folks working in labor mobility.

In this post I want to sketch out some intuitions as to why this is the case.

AI is a general-purpose technology

A general-purpose technology is one that gets used as an input across most of the economy (like steam, electricity, or the computer) so improvements to it ripple into nearly every other sector.

We are in the midst of the development of AI as a general-purpose technology. Many people in the economy are enabling, directly or indirectly, the AI boom. All of those people have the impact of their work amplified across the entire economy, because they are ushering in a technology that will be used by almost everyone in one form or another.

There are many steps in the chain to build and deploy AI. Top AI researchers developing the models. Chip designers and the electrical engineers installing machinery at fabs and data centers. The construction crews and tradespeople building out the grid capacity to power them. The lawyers and lobbyists clearing energy permits. The IT teams at Fortune 500 companies who are implementing AI tools for their companies.

Bottlenecks at any one of these steps can make the entire technology, and therefore the entire economy, less effective1. So the returns to having top talent in roles all across the value chain are higher than they are in normal times.

Fast diffusion + time-sensitivity makes talent even more valuable

Imagine we have two general-purpose technologies undergoing development: Robotic Armadillos and Bison.

• Both are incredibly useful general-purpose robotics technologies that will increase the economy’s output from $20T to $30T once fully developed and deployed.

• Each technology will take 5 years to develop.

• The time it takes for the actual technology to diffuse across the economy varies:

  • Robotic Bison take 40 years to diffuse across the economy, because they are heavy, slow-moving, and resemble untamed agricultural equipment.

  • Robotic Armadillos take only 10 years to diffuse, because they are nimble, cheap, and extremely charismatic.

    

Now: Assume you have the ability to hire a bunch of super talented researchers who can speed up the development of robot technology from 5 to 4 years.

Should these researchers work on Armadillos or Bison?

The short answer: since the full impact of the Armadillos lands sooner, and since we typically discount future value (say at 7%), it is significantly more valuable to put the best researchers on Armadillos rather than Bison.

The math works out to that you get roughly $5T of extra value from speeding Armadillo development by a year, vs. only $2T from speeding Bison.

But perhaps all this math wasn’t quite necessary. The general intuition here is simply: If you have preference for things happening sooner rather than later, then have your talented people to work on technology that will cash out sooner.

The analogy here is: If AI diffuses much more quickly than previous general-purpose technologies (like electricity), then the returns to talent in the development of AI are higher than they were in previous GPT transitions, even if the ultimate economic impact of the two technologies is the same.

Deploy talent into regulatory bottlenecks

Talent is not only increasingly important in the development of general-purpose technologies, it’s also especially important in sectors where there are particular bottlenecks holding back progress.

You genuinely want to deploy talent where there are bottlenecks. And these bottlenecks don’t necessarily have to be technical or economic: They may also be regulatory or systematic.

To pick a couple of favorite causes in the progress studies space:

• Medicine is gated by the regulatory apparatus around clinical trials2.

• Housing is gated by zoning rather than structural engineering3.

This means that to meaningfully improve medicine and housing in society, the marginal top-talented person shouldn’t go build molecules or houses per se; rather they should focus their talent on movement-building, navigating bureaucracy, and getting the rules changed.

The societal returns to extreme talent in navigating these fields are enormous, because the obstacles being moved are what’s holding back enormous public goods.

Conclusion

Inter-company and international competition for technical talent are an outcome of talent’s value during general-purpose technology development. And we can unlock massive public goods by deploying talent towards regulatory bottlenecks.

Ultimately I find this is an incredibly affirming and exciting way to think about the problems we face as humanity. We have so many of them, and we should be absolutely unleashing talented and hungry people at them.

There has never been a better time to become an ambitious and talented person. And if you are ambitious but not yet talented, then deploy your ambition towards becoming talented!

But compare: A Kenyan nurse or software engineer permanently moving to the US is 5-10xing their income for the rest of their lives.

I won’t belabor this point. Migration is, by and large, extremely good for migrants1, and international labor mobility is one of the highest-leverage spaces anyone can work in.

But it’s easy to slip into the wrong mindset here: You probably don’t want to build a solution for the migrants themselves.

Don’t build for migrants

There’s no shortage of migrants who want to move. There is a shortage of companies willing to sponsor visas.

This means if you are building a business to open new pipelines for international mobility, your customer is a company in a rich country.

At the end of the day, there is going to be one employer in a destination country that wants to hire a specific person and sponsor a visa for them. That person making that decision is the person you need to build for2.

Even if you’re the most earnest development entrepreneur who only cares about helping people in poor countries, you still should probably spend most of your time worrying about the person in the rich country.

The migrant is about to 5-10x their income. They’re going to be great. But they’re not going to be be able to move at all if hiring them is not a competitive decision in a foreign market.

Talent export discipline

In How Asia Works, Joe Studwell argues that the defining move of the postwar Asian miracle (Japan, Korea, Taiwan, later China) was forcing domestic companies to compete in global export markets. Not selling to a protected domestic market. Producing for export, to destinations where real customers had real alternatives and would not accept mediocre products out of charity.

Labor mobility works in an analogous way. Labor migration only works when the moving workers are globally competitive. Not “competitive for someone willing to be generous,” but competitive in an actual labor market where the employer has other options. Which means the missing link is almost always destination-specific: a deep understanding of the hiring dynamics, the certification regime, and the employer needs in the specific country you’re trying to route people into.

In global talent mobility, we need people who can make those linkages. People who understand the destination-side customers extremely well, and can hack through the bureaucratic issues needed to get people onto planes.

For this reason, I predict that the successful builders in this space will look like single-destination specialists. They will focus on a specific subset of buyers, learn them really well, and only later think about expanding. The recent initiatives that seem to be working (TERN, Taldo, and Lighthouse of course) look like this.

Remember who your real customer is

If what you’re after is to increase the flow of people across borders, you need to convince people where the grass is greener3.

Focus on that, with maniacal intensity. If you build something that makes a destination-country stakeholder’s life obviously, materially better, the rest of the system will bend to let people through. If you don’t, no amount of supply-side earnestness will fix it.

We want to be healthy. We used to rely on folk remedies and prayers. Now we have safe drinking water and antibiotics, and we live longer than ever.

We want to understand the world. We used to rely on myth and speculation. Now we have telescopes to probe the origins of the universe, and particle accelerators to investigate the fundamental fabric of reality.

We want to love our neighbors, as Christ commands us to. For most of human history, the best you could do was to help someone nearby, or maybe pray for someone far away. But now we can do far better.

Those of us in rich countries have tools to express love of neighbor that our ancestors couldn’t have dreamed of. To take just one example - you can send money directly to people across the world.

If you’re interested shine God’s love into the world as brightly as possible, using all the intellectual and technological tools the modern world puts at your disposal, then you should check out the new book All the Lives You Can Change1.

The core of the book is to ground the ideas of Effective Altruism in Christian teaching2 - and it does a terrific job. In my opinion, the best audience for this book is young Christians: People who who are alight with fire to love their neighbor and improve this broken world, but unsure of the direction they should take.

This book gives you a blueprint, rooted in Christian principles, for thinking about career choice, donations, and volunteering. It’s full of examples of inspiring people who have thought hard about what it means to “love your neighbor as yourself” in the modern world. And in some cases made enormous sacrifices to do so.

C.S. Lewis said that God wants us to have “a child’s heart, but a grown-up’s head.”

He wants us to be simple, single-minded, affectionate, and teachable, as good children are; but He also wants every bit of intelligence we have to be alert at its job, and in first-class fighting trim. The fact that you are giving money to a charity does not mean that you need not try to find out whether that charity is a fraud or not.3

Effective Altruism is, ultimately, a project about how to pursue idealistic goals without being naive.

It’s the best framework for people thinking seriously about serving their neighbors. And All the Lives You Can Change is the best introduction to the topic for Christians.

Highly recommended!

The rollout was a mess. The proclamation dropped late on a Friday afternoon, and different government agencies scrambled to issue different interpretations of the rule (sometimes by tweet). It was initially unclear whether the fee would apply retroactively to everyone currently on H-1B, or only to new applicants. Many employers instructed their employees not to leave the country.

To some extent, the confusion is part of the point2. Trump wants to discourage people from coming to the U.S., and uncertainty achieves this goal.

But the Trump Administration also had a straightforward and uncontroversial goal here: To fix widely-acknowledged issues with the H-1B program3. In this - perhaps surprisingly - Trump also succeeded.

Problems with the H-1B

The H-1B has several extremely well-attested problems, which frankly previous administrations (including Trump 1.0) have done next to nothing about. Two major problems:

• Outsourcing exploitation: The structure of the program is widely seen as easy for large outsourcing/staffing companies to game (e.g., by flooding registrations). This shifts the program away from “best-and-brightest” hiring and toward labor-arbitrage staffing models that undercut wages for Americans. In some years, something like 20% of the H-1B lottery slots go to these outsourcing firms.

• Wage protections don’t work as intended: In theory, the H-1B requires employers to pay enough that hiring H-1B workers shouldn’t undercut comparable U.S. workers. But in prevailing-wage rules, job classification choices, location/wage-level selection, and the worker’s limited ability to switch employers can still allow effective underpayment and weaker bargaining power—particularly in the outsourcing-heavy segment of the program.

Trump decided to take a sledgehammer to these problems ahead of this year’s lottery.

Trump’s fixes: $100k fee and wage-weighting

The Trump administration made two key changes to the H-1B lottery this year:

• The $100,000 fee. This is a prohibitively high fee for most employers, but it only applies to consular petitions. Roughly speaking this means:

  • It applies to people who are outside the U.S. and want to use the H-1B as their initial visa to get into the U.S. Importantly: This is the category that most outsourcing firms fall into.

  • It does not apply to people who are already in the U.S. (e.g., on a student visa).

• A weighted lottery that favors higher-wage jobs. Historically, every entrant into the lottery had an equal chance of getting selected4. This was the first year of a kind of wage-based weighting, where those who are earning higher wages have better odds of getting selected.

Both of these changes are conceptually pretty ugly.

As mentioned above, the $100k fee rollout was confusing. It’s not clear the $100k fee is even legal5. And it serves as essentially a blanket ban on H-1Bs for people not already in the US, which is an incredibly imprecise way to target outsourcing companies.

Wage-weighting sounds great in theory6. But the details matter a lot, and Trump’s immigration policy has not been very detail-oriented, to say the least. The actual method used to implement wage-weighting actually advantages outsourcing firms7.

The impact: So crazy it actually works?

So the changes are undoubtedly clumsy and imprecise policy. But the translation of economics into policy is always clumsy and imprecise. The question is, does the policy move us towards our desired goals?

We won’t get the full answer until national H-1B lottery data is released later this year. But the initial findings:

• There seem to have been roughly 30% fewer lottery entries than previous years. Notably, that’s close to the share of cap-subject filings that historically go through consular processing. Essentially it seems like employers (like outsourcers) who wanted to bring people directly in from outside the US simply gave up because they didn’t want to pay the $100k fee. The fee filtered out roughly who it was supposed to.

• Among applicants already in the US, the wage weighting meaningfully shifted lottery selections toward higher-paying companies. At Lighthouse our clients saw 50% selection rates in the H-1B lottery - and as high as 67% for the highest-paying companies. Historically, odds have been closer to 30%. The most innovative tech companies benefited from the changes to the lottery.

The net effect: The $100K fee that blocked out outsourcing firms. This is a major positive result. And as a secondary benefit, wage-weighting the lottery shifted selections towards the highest-value workers.

Seems pretty good, actually! And it’s worth re-iterating that over the past 20 years, Republican and Democratic congresses, and the administrations of Obama, Trump 1.0, and Biden all failed to move the needle significantly on the improving H-1B lottery. This time around Trump took a swing and hit, if not a home run, at least a single.

Real negative impact: Universities and hospitals

The positive impact is offset to some extent by negative impact on universities and research hospitals.

These employers are typically exempt from the lottery, able to apply for H-1B visas for any qualified candidate at any time of year. The new $100k fee applies to these employers, exactly as it does to for-profit employers.

This means the hospital at the University of Iowa would have to shell out $100k if they want to hire a pediatric oncologist from Toronto8. Rural hospitals already can’t staff their wards, so restricting the supply of international medical talent is a totally unforced error9. And do we really want to tell universities that they can’t hire the best researchers in the world?

As it happens, the $100k fee proclamation actually offers a provision that seemingly would address the university and research hospital issue: The Secretary of Homeland Security has discretion to waive the $100k fee for workers, companies, or industries deemed to be in the national interest. As of now, it seems no such waivers have been issued.

What’s next?

More changes are yet to come. The $100k fee has been set to only be in effect for one year, and government departments are considering various proposals for more permanent solutions.

One change being considered by the Department of Labor would replace the wage-level system with experience benchmarking. This would, in IFP’s analysis, “permanently end employers’ ability to get visas for foreign workers who they will pay less than similarly qualified Americans.” This would be a major step forward.

But this year was a decent start all things considered. We can’t let the perfect be the enemy of the good. In trying to improve the H-1B lottery, the Trump administration moved fast, broke things, and — on net — improved how it works. We can build from here.

—

Thanks to Alexander Kustov and Jeremy Neufeld for comments on an initial draft. All opinions and mistakes my own, not my employer’s.

The foundations of quantum computing as a field were only developed in the 1980s. At that point, the basics of classical computation and quantum mechanics had been understood for decades.

So why wasn’t quantum computing developed earlier?

Nielsen’s account is that while the 70s and 80s didn’t have any fundamental theoretical breakthroughs necessary for quantum computing, there was a shift in the salience of the relevant ideas.

Computation became far more salient in the late '70s and early '80s. It just became a thing which many more people were interested in, partially for very banal reasons. You could go and buy a PC. You could buy an Apple II. You could buy a Commodore 64. You could buy all these kinds of things. It became apparent to people that these were very powerful devices, very interesting to think about.

At the same time, in the quantum case, that was also the time of the … ability to trap single ions. Up to that point, we hadn't really had the ability to manipulate single quantum states.

You got these two separate things that for historically contingent reasons had both matured around 1980 or so.

Think about John von Neumann, who died in 1957. If anyone could have invented quantum computing decades earlier, it was him. But he didn’t.

Not because he lacked the knowledge of theory to be able to do so (he was a pioneer in both quantum mechanics and the theory of computation). But because the potential application of quantum mechanical principles to computing wouldn’t have been very obvious. They only became relatively obvious to researchers in the 80s who had seen the decades of progress in computing technology and our control of quantum systems.

If John von Neumann didn’t see the potential of quantum computing in his time, then there are certainly low-hanging ideas in your field that nobody is pursuing.

So much hasn’t been done simply because no one has bothered to do it. The people best positioned to do it might have other priorities for reasons that have nothing to do with the potential fruitfulness of the area of inquiry.

Become the kind of person who can find the low-hanging fruit. Then find it, and do it.

Competition against a common enemy sharpens focus, motivates hard work, and binds disparate people together.

Races between nations have always driven technological progress and nation-building. Early industrialization was fueled by international rivalries and Great Humiliations. Wars led to radar and penicillin, and the cold war led to both nuclear bombs and nuclear energy.

We want the benefits of international competition without the horrors of war. So rather than channeling our competitive energies into arms buildups and combat, I propose that the US kick off a series of Great Power Races: Competitive, ambitious, positive-sum competitions with other countries.

What makes for a good Great Power Race?

1) An ambitious project, that will take years and require mobilization of a significant part of the economy to meet. The entire nation should be bought in.

2) It has to sound cool. Whether that’s because there’s obvious upside (e.g., curing a disease) or because the outcome itself is just awesome (going to Mars)1.

3) Conceptually clear win conditions. First to achieve X, eliminate Y, or deploy Z technology. A country needs to be able to credibly claim victory at a certain point2.

The heuristic here is: If Trump can tweet about the race, and even people who don’t like Trump would say “yeah it would be cool if we did that”, then it’s a good race.

Some Potential Great Power Races

We’ve only had a handful of true Great Power Races in history. The Space Race worked. You could potentially count the development of COVID vaccines. The Human Genome Project vs. Celera was a race, though not between countries.

But there are tons of projects we could get fired up about and pursue as races. Some examples:

• Curing Alzheimers

• Racing to get a human on Mars

• Creating a nuclear fusion power plant

• Drilling down to Earth’s mantle

• Carrying the post commercial passengers on supersonic flights by 2030

• Creating the first space elevator

The important thing would be framing these ambitious projects in the language of competition (“we’re going to do this BEFORE China/Russia/Europe, because our country is the best!”).

Case study: Curing Alzheimers

Imagine tomorrow, Trump makes a post:

The NIH starts institutes more funding for Alzheimers research, and the FDA creates a framework for trials at scale and accelerated approval pathways. Prizes are offered for specific milestones, visas for top neuroscientists are fast-tracked. States update their science curriculum to emphasize how great the US is at curing Alzheimers, and the White House starts pumping out propaganda materials.

Chinese state media then responds:

While some countries talk loudly, China works diligently. Americans have forgotten the value of discipline and education. We have not. China will cure Alzheimer’s first—for our people and for the world.

The CCP proceeds to make offensive videos of forgetful and overweight US scientists studying Alzheimers, and fills their public billboards with inspirational videos of young Chinese students studying brain scans of patients with dementia.

Over the course of 3 years, both countries announce big milestones in terms of funding, scientific breakthroughs, clinical results. In year China conclusively comes up with a treatment that halts the progression of dementia. The next year, the US creates a pill that prevents Alzheimers altogether. Both countries say that they won the race, Americans and Chinese are both proud of their superior systems, and the whole world benefits from new disease treatments.

But what are they good for?

Kicking off Great Power Races would benefit the US in a few different ways:

• National purpose: The language of competition would foster a sense of pride in country, and sense of contributing to something bigger than yourself (even if just through your tax dollars).

• Locking in commitment to important projects: By publicly committing to defeating other countries in an arbitrary contest, leaders of countries would be locked in to putting their money where their mouth is.

• Shared urgency: Fear of losing to a rival is a great excuse to cut through bureaucratic red tape related to, e.g., permitting, clinical trials, science funding.

• Positive-sum outlet for nationalism: Perhaps by allowing nations to compete on one dimension, they will be less likely to compete in violent ways.

The last point is quite speculative. But since it seems unlikely that international rivalries will ever totally go away, it feels well worth it to explore avenues like Great Power Races that will harness these rivalries for good, rather than ill.

that humans have been on the earth. And still

we’ve only just begun to understand

some basic features of our universe.

In fairness to our ancestors, they were

mostly concerned with finding food and warmth.

Our learning has accelerated since

the scientific revolution when

some men with means and curiosity

developed the techniques to grow our knowledge

which drove explosive growth in industry

and knowledge over the past two centuries.

But still there is so much for us to know.

We use the words “dark energy, dark

matter” to refer to ninety-five

percent of all mass in the universe.

This is code for “we don’t really know

what a majority of existence is made of.”

One of the biggest fundamental questions

that many of us ponder looking up

at night is “are there others out there like us?”

There is not a soul on earth who knows.

Even basic questions I want answered,

like “Did Tyrannosaurus Rex have feathers?”

or “Is it true that dolphins were the vectors

who brought tuberculosis to Native

Americans centuries before Columbus?”

cannot be answered yet.

Three hundred thousand years lived in the dark.

Four hundred years of scientific method.

We’ve only just begun to venture out

into the light.

In Gifts of Athena he makes the case that around 1800 there was a phase change in our ability to generate “useful knowledge”1.

I think it’s quite a compelling take. In this post I try to make this argument digestible.

1.

There are two types of knowledge:

• Ω = propositional knowledge. In everyday language, we call this type of knowledge “facts”. For example “fire is hot”, “iron comes from the ground”, “the earth is a sphere.”

• λ = prescriptive knowledge. This is the knowledge of “how”, these are techniques and technology that get things done. The ability to rub sticks together to make fire, to smelt iron ore, or to navigate using the stars are all examples of λ-knowledge.2

2.

λ-knowledge is often based on Ω-knowledge. Techniques can be generalized more effectively when grounded in understanding of underlying principles.

But any technique’s basis in propositional knowledge can be narrow (based on only very tenuous understanding) or wide (based on extensive knowledge)3.

3.

λ-knowledge can also create new Ω-knowledge. Practical technology generates new understanding of the world.

4.

This is great! Since learning more facts (Ω) lets you build more technology (λ), and new technologies let us discover more facts, then we should be able to create a self-sustaining positive feedback loop.

5.

A self-sustaining feedback loop between Ω and λ is not automatic.

Medieval and ancient peoples invented things all the time, but we did not see rapid and sustained economic and technological growth until the 1800s. Discoveries don’t always kick off consistent, self-sustaining growth4.

This is because societal conditions can hinder the ability of Ω and λ to generate each other.

• Lack of consensus on grounds for knowledge: If people don’t agree on methods of producing knowledge (e.g., if some think knowledge comes from experience while others think it must come from divine revelation), Ω-knowledge won’t be accepted or spread.

• Lost knowledge: If knowledge isn’t written down or codified, it can die with its owner.

• Lack of access: If access to useful knowledge is limited (e.g., through limited literacy), it won’t diffuse broadly.

• Norms: If there aren’t norms and resources incentivizing understanding why a technique works, λ-knowledge stays narrow and disconnected from Ω.

6.

The industrial revolution was a result of a phase change in the generation of useful knowledge.

In the centuries preceding the industrial revolution, the right institutions and culture developed, allowing the feedback loop between Ω and λ finally became self-sustaining in the late 1700s.

Mokyr emphasizes a few forces here:

• Epistemic consensus: Scientific societies established shared standards for how knowledge is produced and validated, so Ω-knowledge could spread with authority.

• Knowledge preservation through institutions: Universities, publishers, and laboratories gave knowledge a permanent home, and the “Republic of Letters” ensured that learnings were shared across borders.

• Access to useful knowledge: Encyclopedias, printed manuals, public lectures, and correspondence networks dramatically lowered the cost of accessing both Ω and λ knowledge.

• Industrial Enlightenment Norms: A cultural shift in which educated elites began taking technology seriously as a subject of scientific inquiry, creating norms that incentivized understanding why techniques work.

Conclusion

We are the beneficiaries of this feedback loop. The accumulation of Ω and λ - continues today, and is the basis for our absurdly high material standard of living.

It’s up to us to keep it up. When thinking about every institution, norm, or policy, we should ask the question: “Does this make it easier or harder for knowledge to compound?”

In the long run, few things matter more.

• Some this is due to institutions.

• Some this is due to culture.

Obviously these things are related to each other.

It is clear is that population-level genetics is not a major driver of skill differences in the vast majority of fields.

Which is great, because while you can’t change your genes, you can opt into a new culture, or seek out new institutions under which to flourish.

And it also clear that the downsides of discriminating against people on the basis of genetics are so insidious, that it is worth having a strong cultural taboo against this kind of discrimination.

Several of those arguing that Anthropic should submit made reference to the thinker Carl Schmitt.

I had never heard of Carl Schmitt until yesterday. He was a Germany political theorist who joined the Nazi party. Per Wikipedia, “Schmitt's doctrine helped clear the way for Hitler's rise to power by providing the theoretical legal foundation of the Nazi regime.”2

It seems like Schmitt was an important thinker. Some of his ideas—like politics as fundamentally friend vs. enemy—seem useful for explaining contemporary political dynamics. And in general it can be valuable to grapple with strong systemic thinkers, even if you disagree with them.

But if you are engaging in a debate…

…trying to persuade others to your point of view…

…and the strongest thinker you can muster to your side is a Nazi…

…then you might want to reconsider the merits of your position.

I’ve got a reverse Uno card for you here. My Chinese mother-in-law is blogging about trips she’s taken to the US over the past decade.

Her blog series is called 说说我心中的美国, which translates to something like “America from my perspective”. Here are of my favorite excerpts from her observations2.

“America’s declining industrial heartland”

Hilariously blunt observations throughout.

New York is “a tired city, with dirty streets, dark subways… and different psychopaths you encounter every time you go outside”.

She thinks her daughter’s roommates are “academically stressed and physically ill” and “probably a little autistic.”

But her harshest words are for 2014 Berkshire County Massachusetts.

In the evenings, we stayed in Berkshire, a dozen kilometers away from our school, to experience America’s declining industrial heartland. The county seat should have a lot of residents, with hotels, restaurants and supermarkets, but the whole city doesn’t have much energy, and it looks very dead. At some of the former factories, the walls were filled with beautiful graffiti, but the interior was empty. My daughter told me that this was the city’s effort to save the old town with art. She appreciated it, but now it seemed that these measures would not be able to revive an old American town.

When a Berkshire county official offered to help her find parking, she took this as a sign that the county was in desperate need of tourists.

Nice Americans and cold courtesy

Her overall impression of Americans is quite positive.

That was my sister’s first trip to the United States. Throughout the trip, she was repeatedly amazed by the courtesy and warmth of the Americans… I had also been treated so kindly: at the New York airport, the girl who took me directly to my destination after I asked for directions; at the airport, the ground staff who greeted me in Chinese and pointed me to my gate; at the hotel, the staff member who repeatedly helped me with my itinerary for the next day; the B&B owner who drove me to explore the countryside...

…We went to America with preconceived notions, thinking it was an arrogant country…

America is nothing special; the people of that country are no different from the Chinese—most are friendly and trustworthy. We really can’t label a country or a people based on a single quality.

Heartwarming. Pretty standard observation when visiting a new country. “Maybe we aren’t so different, you and I.”

More to my surprise, was that she specifically praised the “efficiency and flexibility of the Americans”.

She describes my family as “comfortable, stress-free, and having slightly cold courtesy”. Basically, “midwest nice through a language barrier.”

“No need to worry about a collision”

She was quite impressed with the orderliness of US roads.

Although we can clearly see the vehicles coming across the road, the barrier in front of the middle is open and there is no need to worry about a collision.

In a similar vein:

I spent three days in Williams. What impressed me most during those three days was that whenever we stood at one end of a zebra crossing, the cars on the road would stop and wait for us to cross before they continued.

Now, on my travels to China, I have never observed the roads to be particularly disorderly. But my mother-in-law’s baseline was probably set in the much-more-chaotic early 2000s when she learned to drive.

“Chinese culture is creeping into New Yorkers’ lives”

Her first observation of New York is a classic one:

It’s the worst city because it can destroy you instantly to nothing, and the best city because there are so many opportunities.

The garbage is not disposed of on time in her NYC hotel. She is alarmed by several interactions with strangers. And she is impressed out how quickly her daughter knows to walk away when a commotion breaks out on the street.

Heavy blows come for the New York Subway, which is sensible if you’ve ever ridden the clean and well-ordered subways of Suzhou.

The New York subway, in particular, which doesn’t have a bathroom at all, is a terrible experience3.

I love my home, Suzhou, a city where you can close your eyes on the subway.

And she is struck by the Chinese influence in New York.

The Chinese language that can be seen everywhere in New York, the popular Chinese restaurants, the Chinese songs that can be heard on the subway, and the roads that can be asked in Chinese anytime, anywhere, is the true cultural invasion! Chinese culture is creeping into New Yorkers’ lives.

Overall: “Life in New York was tense.”

Same.

Conclusion

It’s fun to see your home and culture reflected through the eyes of others.

These posts have also helped me build my perspective on the value of travel.

It’s easy to make fun of people who take a trip to another country, and then think they understand it at a deep level4.

But traveling definitely teaches you something. I wouldn’t trust my mother-in-law to predict future geopolitics on the basis of her visits, but I do think she has a pretty good feel for what American culture is like, and the ways in which Americans are different from and similar to Chinese.

After reading these blog posts, I’m marginally more bullish on how well you can understand a country by taking a trip or two as a tourist.

Travel more!

My internet history is a similar story of stops and starts, full of half-consumed and mostly-forgotten wikipedia pages, youtube videos, and reddit threads.

I love learning about the past. I’m a great admirer of humanity, and never get bored with trying to wrap my head around the place and time we occupy in the grand sweep of history1.

But I am also incredibly forgetful. After finishing a book, podcast, or article, I rarely retain much. Like a statue of Ozymandias, my knowledge crumbles, and I’m left with nothing but a memory of a name, or a sense that some city was important.

I’ve long envisioned having a visual archive to make my knowledge of history sticky. A vague dream of some kind of an ever-growing timeline into which I could codify my knowledge2.

I’ve tried journaling, using spreadsheets, and various online tools. None have worked well enough.

So since software is free now, I decided to build it myself.

Stone, Paper, Silicon is a personal archive for hobby historians. It is NOT a complete timeline of humanity, but rather a basic and flexible framework into which you can use to slot in historical knowledge as you learn.

It has three main features: Timeline, World Map, and Quiz.

Timeline: This is anchor. You can see the framework of history laid out before you.

You can add timeline entities here for states, religions, cultures, and events you learn about.

World Map: Explore the rise and fall of states, nations, and empires throughout history, year by year. If you see a state you want to learn more about, click on it, and add it as an entity to your timeline.

Quiz: Quiz yourself on historical events that have been added by all users.

—

For now I’ve built very basic multi-user functionality. You simply enter your name when you land on the site. Any timeline entities you create will appear under your profile and will be publicly available to others.

But enough of my words. Give it a try, and let me know what you think!

Between say 800 and 1300, the number of Europeans tripled. And the percent of people living in towns went from 2%→8%.

Much of this social change was driven - if not in full, then at least in significant part - by the development of new technology1.

The diffusion of the heavy plough between 800-1000 allowed for easier farming in climates with heavy, wet soil. But 6-8 oxen were required to pull a heavy plow, which was far more than a typical peasant owned. Higher capital needs helped drive village nucleation and tilling in adjacent strips of land rather than in isolated hamlets.

The adoption of metal horseshoes in the ~900s greatly increased the usefulness of horses, which had a variety of effects. My favorite of these was described as follows:

The ox moved so slowly that peasants using oxen had to live close to their fields. With the employment of the horse both for ploughing and for hauling, the same amount of time spent going to and from the fields would enable the peasant to travel a much greater distance… Even a slight increase in the distance which it was convenient to travel from the village to the farthest field would greatly enlarge the total arable which could be exploited from that village. Thus extensive regions once scattered with tiny hamlets came to be cultivated wilderness dominated by huge villages…which in architecture and even in mode of life became astonishingly urban2.

This horse-driven-agglomeration is particularly striking to me, because it’s like an inverse mirror of what was to come with the automobile: When people got horses they moved in to the villages, but as people got cars they moved out to the suburbs.

Heavy armor and the stirrup were adopted between 700-1000, which helped enable an entirely new kind of horse-mounted soldier: the knight. An appropriately skilled and armored knight required a high level of investment: A lifetime of training, advanced weaponry and armor, the best horses. So “a gulf appeared between a warrior aristocracy and the mass of peasants.”

—

I by no means want to imply that the technologies listed above were the only factors in social changes in the middle ages. Other institutional, religious, climatic, and demographic factors were obviously also at play.

But technology definitely had an important effect, and in ways that can be hard to predict and take years to play out. It’s not obvious ahead of time that horses would contribute to the development of villages, and that the plow would drive collectivization.

And it’s worth keeping in mind that medieval technological (and social) changes were tiny compared to those to come with the industrial revolution.

With that perspective, medieval Europe looks technologically and socially static. But in reality, it wasn’t, and history never is. We never reach equilibrium.

Right now we are in the midst of3 an AI boom, and look to be accelerating faster than ever. It may just be continual technological and social change from here on out. And our descendents in 200 years might live with a pace of change so fast that our own era may appear almost static by comparison.

On the topics it covers, it’s quite thorough. I want to extend the argument to cover some additional points related to high-skilled immigrant assimilation1.

From an economic perspective, it is clear that the US as a whole greatly benefits from high-skill immigration. Talent Maxxing helps the US economy. But immigration is about culture too, and the culture of immigrants have a lot of people worried.

To be clear at the start: I think “we will lose our culture” is a valid type of concern to have. It is not necessarily a sign of bigotry to be worried about losing what is distinctly American.

I lived in Kenya for four years. It is hard to live as an alien in another culture, where people don’t speak your language, share your values, reference the same culture, or celebrate the same holidays.

So I understand why people hesitate at the prospect of their country changing so quickly that it no longer feels like home.

We are lucky that - for the most part - high-skill immigration does not recreate the kind of cultural alienation people worry about2.

Here is the tldr before I go into a bit more detail on each point below3.

If you are worried that Talent Maxxing will lead to too many people who…

• …don’t speak English → don’t be.

• …will not put America first → don’t be.

• …do not share commitment to American values → don’t be.

• …lead to less cultural homogeneity → some validity here.

• …are not white → I’m not sure I have much to say to you.

English

I haven’t found great data on English skills of high-skilled immigrants specifically, but it seems clear it’s quite high. Some back-of-the-envelope math with ChatGPT estimates 90%4.

Patriotism

Surprisingly, there isn’t tons of great data on how patriotic immigrants are. But indications are they are at least as patriotic, if not more, than native-born Americans.

American values

The Hamilton Society debate on high-skilled immigration opened with an argument from the anti-immigration side that immigration “threatens to destroy the foundation of this great nation by injecting massive quantities of alien subjectivities into the very constitution of the American polity.”

A few points here that I believe are worth mentioning:

• Half of tech unicorns are founded by immigrants. This is often trotted out as a economic argument, but it is also a cultural one. Skilled immigrants are often deeply capitalist and committed to innovation.

• There is selection bias in terms of the immigrants who come here. We get the most individualist, ambitious, and capitalist ones. People who love China’s social cohesion or Europe’s social safety net do not typically move to the US.

Cultural homogeneity

One opponent of immigration at the Hamilton Society debate pointed out that recently he has seen fewer Christmas lights and less trick-or-treating, and that this is a result of immigrants who do not celebrate American holidays. In fact, the market for Christmas lights seems to be bigger than ever, and the decline in trick-or-treating seems driven by a variety of factors that have little to do with immigration.

In general, I think we should be skeptical towards arguments of the form “something that I like is disappearing from the culture, and it must be because of the immigrants.”

But to take a more generous view of this argument: Immigration skeptics here are pointing to a sense that a certain kind of cultural cohesion gets lost when new people enter a society. Which is fair to some extent; there have been some studies that back this idea up.

To the extent you think cultural cohesion is extremely important, you might justifiably be worried about having too many immigrants in your neighborhood who don’t assimilate.

But it is also important to keep in mind that people had the same kinds of worries about cultural cohesion 130 years ago when it was southern European Catholics and Eastern European Jews who were immigrating to the US. Many of these groups have assimilated into the US to the extent that they aren’t even thought of as foreign anymore. Culture is malleable5.

Of course there is always…

…bad old-fashioned racism. Not everyone who is worried about culture is racist. But racists almost always do talk about culture.

The racist objection to high-skilled immigration is not one I will cover here.

Conclusion

Just to end this on a classic and hopeful image of the US as a melting pot. Talent Maxxing means getting the most talented people from to bring their ideas, music, food, and traditions here. And then to assimilate and to pursue their work and life in a distinctly American style: optimistic, capitalistic, bombastic, free, and individualistic.

The American values of life, liberty, and the pursuit of happiness are powerful because they are universal.

The economic literature is clear on the benefits of agglomeration. High-skill workers produce more when they cluster in cities, in top universities, and in the best companies. And the benefits of improved productivity spill over to society at large.

This leads to a clear conclusion: The US should be Talent-Maxxing.

We should attempt to collect within our borders all of the most talented scientists and technical minds, in the world1.

Talent-Maxxing means doing two things:

1. Deeply investing into developing American talent. This is something basically everyone agrees we should do in theory. It’s not a topic to which I feel I have much to add, other than to say we certainly can and should do much better than we are now.

2. Acquire all the best non-American talent in the world. Not everyone agrees that we should do this, and this post will focus on arguing that we should. But it should be remembered that #1 is equally if not more important.

So why should the US try to hoover up all the world’s best scientist and technologists? In short, because we can put them to work making life better for Americans and for the world.

Attracting top talent is a big a win for us US citizens. The most talented immigrants:

• Build great companies that benefit the public. Better access to talent improves firm performance, which drives economic growth. The 62% of Americans who own stock are the beneficiaries of this growth.

• Create jobs for Americans. Studies find skilled migration increases employment of native Americans on balance2.

• Contribute wealth towards the US treasury. The average college-educated immigrant contributes on the order of $1M to the US treasury over 30 years.

• Generate positive externalities for all Americans. When an immigrant scientist develops a life-saving drug or an immigrant engineer builds a better product, the benefits spill over to everyone who uses them.

In fact the spillover effects from improved medicine, scientific advancements, and new technology do not end at our borders. When the state of the art advances, the entire world benefits.

Now you might be worried that trying to attract all the most talented people in the world would mean that the rest of the world is left talentless. Don’t worry.

1. The supply of talent is elastic. Talent comes as a result of investment into human capital. Often when people have opportunities to move abroad, they invest more in education, which can have positive talent spillover effects in their country of origin.

2. Other countries will continue compete with us for talent. If the US invested more heavily in obtaining all the world’s top talent, other countries would step up their game. They would improve incentives to immigrate, easing visa barriers, and improve policies. The net result being better global options for the world’s best talent3.

3. The migrants do not always stay in the US forever. The general finding is that “between 30% and 60% of migrants will likely return to their home country within ten years.” When they return, the bring new skills, and US connections that benefit their home country.

This is not even to mention the fiscal benefits that accrue to family and friends in origin countries as a result of remittances (typically 10-15% of immigrant earnings). In general we should not be worried about brain drain.

And of course, making the US more attractive to global talent benefits those talented individuals themselves4.

What would Talent-Maxxing actually look like? Some initial thoughts:

• Be unapologetic about the fact that we want the world’s most talented people to come to the US and become American.

• Reform our immigration system to focus on skills. Much has been written about specific policy proposals here. Wage-ranking the H-1B would be a great start5.

• Build the critical talent infrastructure to allow public and private actors to identify, evaluate, and develop global talent in the US interest.

We are lucky enough in the US to have culture, ideals, and an economy that many around the world aspire to. We have our pick of who we let in, and we should let in the people who will clearly benefit American citizens.

So let their people come, and let them serve out people.

This was an incredibly fun event. It was raucous, it was heated. Someone got kicked out. People tried and failed to follow parliamentary procedure1.

And it was also thought-provoking. Not all arguments put forth were legitimate, and not all were good. But as someone strongly in favor of more high-skilled immigration, I emerge from the evening with a better understanding of why someone would oppose it.

There were several opposing arguments that I think are worth taking very seriously2. To put them in my own words:

1. Obvious problems with the H-1B taint skilled immigration overall. It’s been well-documented that outsourcing companies use the H-1B to essentially replace US workers. If implementation of current skilled visa pathways allows such abuse, then why would we think any future program would be better?

2. Skilled immigration does not exist in isolation, but is a piece of a larger, broken system. You may have heard: There are millions of undocumented immigrants in the US, and people are not happy about this. Emotionally and optically, skilled migration cannot be separated from larger debates on who comes into the country.

3. Immigration is about culture, not just about economics. What kinds of communities do we want to build? What is our shared vision of the future of our great nation? Just showing that immigrants drives GDP growth or firm success won’t cut it.

When I arrived at the event, I had planned to speak, making some points in favor of high-skilled immigration. I was politely told that my arguments weren’t very strong, and after observing the debate, I see why. My points were based on imagined objections to high-skilled migration, not the actual concerns people have.

Now as it happens, I think there are extremely good responses to all of these objections3. They were made with varying degrees of forcefulness by the pro-high-skill-immigration side of the debate.

But the value of debate is having these arguments surfaced and made in their full force before you. Coming face-to-face with the actual argument of the opponent, not the foe you have imagined. Thank you to the Hamilton Society for putting on this debate, and I look forward to attending participating in the next one.

The first is The Metamorphosis of Prime Intellect. In this book, a computer wedded to Asimov’s three laws of robotics becomes God. Its near-unlimited power and single-minded devotion to its laws leads to many horrifying and unintended consequences.

The second is Robert Caro’s four1-volume biography of Lyndon Johnson These are about how LBJ ruthlessly accumulated political power in a corrupt and democratic society. And then he wielded it as he pleased, for good and for ill. Enriching himself, passing the Civil Rights Act, launching the Great Society, and miring the US in Vietnam.

Caro mentions almost offhandedly in an afterward of one of the books that “[LBJ] never went so far as assassination. Because he stayed in the lanes of what is acceptable in democracy.” Nobody in the US has immunity from committing murder. However much power LBJ accumulated, if he had had a political rival assassinated, he would have been caught and convicted. And that would be that.

The startlingly simple take-away for me is that of the importance of checks and balances.

When thinking about the good society, one of our principles worries should always be to avoid giving any person or institution too much power.

A surprising number of apocalyptic scenarios really just boil down to “X will get too much power, then use it to harm us.” This includes AI labs, the US president, the CCP, a global government, or a single AI itself.

Seems like a good case for just sticking to the tenets of classical liberalism. Too much power should not be stored in one place. Not in any individual. Not in any given government, or branch of government. Not in a company or an industry.

No matter how noble, no matter how wise that person or institution may seem.

Power should be checked and balanced.

When I first joined BCG I was excited by the impact I would be able to have. BCG does a lot of work with the Gates Foundation, and I was optimistic that I could get on one of those projects and help solve global poverty or something.

But I quickly realized two things:

1. Those projects were incredibly competitive to get onto. My odds were not good.

2. Even if I did get on a project, I would know that had I not been assigned to it, someone equally-capable would have been on it. If I vanished, the slide deck would still ship, the meeting would still happen, the world would not notice.

I was a cog in a machine, and ultimately extremely replaceable.

This is just one example of a feeling I suspect others may share. A melencholy about the fact that if we were to disappear, our work would go on without us. Though our would be sad of course, overall the world would be unaffected by our disappearance1.

But replaceability is the price we pay for resilience. Human societies are flexible, adaptive systems that do not depend on any one member. If a person disappears, work gets rebalanced and the ship keeps sailing.

Imagine that you really were indispensable to the functioning of your town, or nation. That you were the bolt holding the entire artifice together. How crushing would that feel?

I for one am happy to be redundant rather than indispensable.

It is noble to be part of a flexible system - where none of us individually are ultimately mission-critical, but where all of us collectively can achieve great things.

1. The midpoint of history is 1300. And 10% of history has occurred since 2000.

2. Historically new technologies have taken decades to diffuse throughout the economy. AI will be the same1.

3. Politicians are not attuned to policy details in immigration. And these details matter2. This issue is not unique to immigration.

4. The lesson of history is to have low expectations.

5. I am so much better off than a lot of people. And there is no moral justification for this state of affairs.

6. Your solution to any given problem probably isn’t the right way, and definitely isn’t the only way. Constantly, constantly expose yourself to new ideas, and new people who will push you. Relatedly, if you aren’t talking to your customers, you aren’t learning.

About myself:

1. I’ve had a step change in my quality of life since getting married. Would recommend!

2. My semi-constant sense of hurry is self-imposed, and an obstacle to getting actual work done, and to being warm and outgoingness.

3. I no longer optimize for impact as much as I used to. It’s not clear it was that effective anyways.

4. I am not nearly as other-focused as I wish to be.

This post is me trying to rationalize why I and many others spend so much time and energy on paperwork that doesn’t directly produce anything of value.

It’s a commonly-observed and annoying part of modern life that we have so much paperwork. But an inevitable part of having a social contract of any complexity is that some things have to be spelled out in contractual terms. This means you need lawyers, analysts, and inspectors ensuring that contracts are being followed, the i’s dotted and t’s crossed.

Paperwork, generally, produces compliance. Industrial safety compliance, HR compliance, compliance with financial, employment, and immigration laws. And compliance at a company level leads to trust at a systemic level.

Right now things are as good as they have ever been in large part because we can live without constant fear of being poisoned, exploited, or defrauded by the companies we trust with our money. I can in general, trust that a set of rules are being followed.

In my field specifically, a well-functioning US immigration and border system builds trust with the public that the 1 million people who enter the US each day do so with legitimate reason, and in an orderly fashion. Immigrants have to know that when they arrive in the US, they will be accorded some basic rights. And employers need to know that they can hire foreigners without fear of arbitrary deportation1.

Trust is an essential output of the immigration system. And in general, trust-producing industries are crucial to our society’s functioning.

But we should not tolerate compliance for compliance’s sake. Systemic trust, not compliance, is the end goal. And trust, like any good or service, can be produced efficiently or inefficiently. I for one, do not think we should be required to physically print out and mail in visa applications2.