What You Can't Hold

October 2023. Israel announces the largest reserve mobilization since 1973: 360,000 people. A call, a message, an order. Tens of thousands of Israelis are abroad at that moment. A honeymoon in Thailand, a medical student in Lithuania, a tech worker in New York. None of them is required to return. No law reaches them there, no contract. They board planes. The newlywed couple turns their tickets around and flies back. The student puts down his textbooks, because his friends are already at war and reading is impossible. The New York worker returns to his old medical unit. In WhatsApp groups across ten countries, people without call‑up papers ask to be summoned. El Al is the only airline still flying. Reuters carries the words of one of them: “Everyone is coming back. Nobody says no.”

That is precisely what interests me. The mechanics of what holds a person when no external force presses on them. What makes them turn a plane around and fly toward danger. The moment every step needs to be explained, the program’s frame becomes too narrow. And as soon as one person breaks the condition, the whole thing collapses, because it rested on nothing but a promise. The mechanism contains no coercion, so what is written holds exactly as long as everyone agrees. It is something a person carries inside. And before asking how to cultivate that something, you have to understand what it stands on.

In 1979, six years after the war against which the scale of the mobilization is measured, two Hebrew University professors, Felix Dothan and Shaul Yatziv, came to Chief of Staff Rafael Eitan with an idea. The idea was simple and painful: the army was tearing the best minds away from science, science was not giving the army what it needed, and the gap was being paid for in lives. They proposed a program that would take eighteen year olds, teach them physics, mathematics, and computer science at the university for forty months in Air Force uniform, with field training in every branch, and then keep them in the system for six years. Nine years in total. The bet that “creative abilities peak early” drove the launch of the program on that very belief. They named it Talpiot after the tower in the Song of Songs. The aim was practical: to grow engineers who would stay inside.

The program was built on what already existed. Military service in Israel is thirty‑two months for men, and that is since 2015; under the 2008 law, reservists are called up until forty, officers until forty‑five. The man on his honeymoon was a reservist. The program gave him neither knowledge nor network; he was never its cadet. What made him return was older than the program. Older than the state. It was on that layer that they built.

In April 2026, the press reports on a study by the Stanford Graduate School of Business dedicated to the program. The numbers: roughly three percent of Talpiot graduates founded companies worth over a billion dollars. For comparison, among Stanford MBA graduates the figure is about six per thousand, meaning 0.6 percent. Nearly a quarter of the program’s graduates became entrepreneurs; they account for around ten percent of Israeli unicorns and thirty percent of deca unicorns. A result no one had set as a goal.

Wiz. Founded in 2020, bought by Alphabet for 32 billion dollars, the deal closed on March 11, 2026, as the largest exit in the history of Israeli tech. Three of the four founders are Talpiot graduates. The buyer sits in California.

The Wall Street Journal reported a different yardstick in 2007: “In its goal of creating a new generation of military leaders, critics say, the program has not been a success.” Military leaders, that is, officers. Two yardsticks stand side by side: one measures what the creators intended, the other what no one planned. The measured product of the program is those who came out of it.

The state looks at this product and sees a leak. In 2023, 86,509 people left Israel; in 2024, 91,499 left; in 2025, 90,922 left. Roughly half the startups founded in 2025 are registered abroad. The Tax Authority is examining a tax on startups launched by graduates of elite technical units, to apply “up to ten years after service, even if the companies are registered abroad.” The state’s argument: the army invested resources, like a university in a spin‑off, and is entitled to royalties. The objections are printed in the same article: tax treaties with the United States, freedom of employment, an incentive to serve elsewhere. The state reaches out for what is slipping away, and hits an answer.

Albert Hirschman described this mechanism in 1970 in his book Exit, Voice, and Loyalty: when quality drops, the consumer changes supplier. Voice cast with one’s feet is exit. Loyalty gives voice time to be heard before the person leaves. But loyalty comes in more than one kind. There is loyalty to a contract, signed and torn up. And there is the kind a person carries inside.

“Next year in Jerusalem.” The phrase is spoken at the end of the Passover Seder and at the close of Yom Kippur. It is not biblical; it is first attested in the Middle Ages. Nineteen centuries without a state, without an army, without a border. No one to compel, no one to punish non‑compliance. Yet the phrase lives, handed from parents to children, travels with a person anywhere, and holds the direction. Herman Melville, in his 1876 poem Clarel, read it as a split loyalty: a man in New York looks east, torn between two worlds. He got the geography right, but the tearing wrong. A person who carries this phrase inside is not torn. They carry the center with them.

This is the formation. An order a person carries within, existing not as a prescription, not as a contract, not as elite membership. It is not granted with a diploma, not instituted by law. It is either there or it is not. And if it is there, it works where no state machinery reaches: in Thailand, in Lithuania, in New York. The Talpiot program did not create this formation; it received it ready‑made and built nine years of knowledge, network, and skill on top. The network that journalist Jason Gewirtz described in his 2016 book as “one degree of separation from any answer, two degrees from funding” is not the program’s invention either. It existed before the program, without walls and without a charter. The program gave it tools and shared experience.

AnnaLee Saxenian showed in 2002 how “brain drain” becomes “brain circulation.” Engineers from Taiwan, India, and Israel who went to Silicon Valley return with networks, capital, reforms. Israel’s venture capital industry of the late 1980s and early 1990s grew precisely out of such returns: repatriate engineers together with state policy. The Taub Center, in its 2025 report, phrases it cautiously: emigration is “a loss, at least in the short term, though in the long term it may have network advantages.” The outcome depends on who returns. And to what.

The condition for return is the formation. The network returns to the foundation. Capital returns to the foundation. If there is no foundation, a person leaves and does not come back. The state can take all the offense it wants, call it ingratitude, send a bill. But a bill is sent to someone who took the goods. Someone who carries the formation within took nothing on credit.

People are retreating more and more into rationalism these days, ungrateful for the time spent on them. That is the builder’s pain. The pain of one who invested years in a system meant to hold, and received instead a calculation: graduates weigh what is more advantageous and go where it is warmer. Ingratitude serves as a symptom. A symptom that the foundation was never there to begin with. That a person selected for ability and trained by the program never received what is older than the program. The thing that holds without a contract.

Jews, at least, more or less hold the national formation wherever they are. This is an observation about an institution that has worked for nineteen centuries without a state apparatus. The Seder, Yom Kippur, a phrase spoken aloud once a year. Habit. Discipline. The Talpiot program selected those who already had the formation and gave them nine years of a shared path. Nine years on top of nineteen centuries.

A nation must be a system that holds when others break. That is why I am a nationalist. I am honestly not one of those who build a cult of the nation around embroidered shirts and slogans. What matters to me is the mechanics: what holds a person when nothing presses on them. What makes them turn a plane around. The formation does not work like a program. It cannot be introduced by decree, cannot be measured by a test, cannot be instilled over nine years of schooling. Such tools are powerless to establish it. And if you have it, you do not count the cost.

Because the one who carries the formation does not ask whether it pays. They board the plane. They do not weigh whether the state is grateful or not. They do what they are meant to do, because they cannot do otherwise. And when the state tries to collect a tax from them for becoming what they became, it does not understand whom it is dealing with. It thinks it is dealing with a consumer. It is dealing with a bearer.

Beyond this text lie another country, another war. 5.9 million refugees from Ukraine abroad, February 2026. The share of those who want to return is falling: 65 percent at the start of 2024, 61 percent in November, 57 percent in the latest survey. 4.2 million have returned, a million of whom came from abroad. The numbers are silent about what exactly brought these people back. But the question the builder faces does not disappear: is there any sense in building what you cannot hold.

The answer begins with what a person will carry with them when they step beyond your reach. What they will tell themselves at night when no one is listening. What will make them turn the plane around. If that is absent, you will not hold them. No tax will help, no contract, no elite network. The person will go where it is more advantageous and will be right by Hirschman: they are a consumer, after all, they vote with their feet.

But if it is there, then you can build. Select those who already carry it. Give them knowledge, shared experience, nine years of the path. The network will grow on its own. The state will get engineers who build 32‑billion‑dollar companies, and reservists who return from their honeymoon to take their place in the ranks. But first, the foundation is needed. First, what is older than the program, older than the state, older than any calculation is needed. First, the formation is needed.

Everything else is a superstructure. A superstructure holds exactly as long as the foundation holds. And when you, the builder, look at your own project, the school, the unit, the system you are creating for your country, ask yourself: what of all this will a person carry with them when they walk out the door.