The Trade School 2.0
America needs.

Zane Hengsperger's recent thread on reindustrialization has been circulating for a reason. He laid out eight bets a founder should take on America's industrial future, and the eighth — America's Trade School — is the one that makes the other seven possible.

You cannot build reshored factories, defense primes, semiconductor fabs, or energy infrastructure without people who know how to weld, machine, fab, maintain, and operate. And those people are disappearing faster than the country can replace them.

His diagnosis is exactly right. This is the bet that unlocks every other bet.

And I am here to share what exactly we are building to tackle that in partnership with some of the most important companies in America.

What we are building: Trade School 2.0

Trade School 2.0 is a new kind of institution. It is not a school attached to a factory. It is a factory with a school inside it.

The whole bet is collapsing the gap between learning and doing. Universities leave students sitting in classrooms that never touch the work. Trade schools run outdated curricula on outdated equipment. Neither produces what the market requires.

The model is three moves:

1. Acquire a working manufacturing business — one with real contracts, real customers, and real revenue. We buy instead of build because an existing shop gives us the one thing no classroom can: real work, already happening, that students step into on day one. Plus the existing talent.
2. Automate it with Arklight OS — our unified operating system for the floor and the school. It runs the ERP, instruments every machine and contract, and feeds the training program with real-time data on how work is actually being done. Currently being battle tested.
3. Embed a cohort of students inside it — paid to train on live defense and commercial contracts from week one, mentored by master tradesmen, measured in real time by the software running the floor.

The school and the factory are the same place. The work, the stake, and the consequences are real.

The business model

Our bet is that the revenue will come from four places: employers, the factory itself, the federal/state government, and — eventually — modest tuition recovered against performance. The employer pays because the alternative is a nine-month open role at $150K+ in lost output. The government pays because the alternative is a stalled CHIPS fab. We're designing the institution to pay students on a pro-rata, performance basis — with the goal of offsetting their full tuition and more by the time they graduate.

We are betting on kids other institutions have written off. Joe Liemandt is proving at Alpha School ATX that two-hour learning beats six-hour sitting. The 1517 Fund (William Blake, Danielle Strachman, Zak Slayback) has proven time and again that the most talented young people in America are worth supporting and backing. Alex Karp and Shyam Sankar built Palantir on the idea that hard problems deserve serious young people.

For far too long, this country has underestimated its eighteen-year-olds. Give one of them the right profile, a real contract, a mentor to learn from, and stakes that matter, and you transform them into the most elite industrial operator in the country.

That is the bet.

Why this needs to exist

Meanwhile China has 34 million students in vocational education and supplies 70% of the new skilled workers entering Chinese industrial industry.

Germany and Switzerland out-structured us. And China has militarized their education system, which you can find how they did it here — it is scary.

Why I'm building this

I spent over a year on the board of an university's foundation. I tried to reform it from the inside. The governance structure made meaningful change structurally impossible. California made it legally impossible. Every reform I proposed was shot dead a mile away.

That experience convinced me of one thing: the fix cannot come from inside the system. It has to be built outside it.

Last year I shared this with Trae Stephens. He introduced us to Anduril. Anduril led to Hadrian, to the U.S. Navy's Maritime Industrial Base Program. We had a hunch that there should be a new type of institution — one that partners directly with the critical industries to develop a sustainable, predictable and elite pipeline for the national security industrial base. Think a privatized West Point, or the Battle School from Ender's Game, but for reindustrialization.

Every conversation confirmed and sharpened the thesis. Every serious defense and advanced-manufacturing company in the country is hitting the same wall. All of them have experienced the pain of a talent system that cannot scale to meet the moment.

So, I left the university, and my AI job to tackle this.

The not-so-controversial pedagogy

It's easy to talk about the Arklight OS, or the factory. But what makes us different from any other existing institution is our pedagogy.

We looked at where humans actually learn fastest — and it isn't Harvard, Stanford, community college, or some fancy trade school. It's the military.

These institutions (e.g., Nuclear School, Language Institute) figured out decades ago what higher education has forgotten: humans learn when failure has consequences, when application is immediate, and when stress is gradually inoculated.

Most education efforts fail the opposite way. They speed up content delivery and call it learning. Content delivery is not learning, failure is.

We built the 'Arklight Pedagogy' on how the military trains for its hardest jobs — where errors cost millions, lives are on the line, and the timeline is months, not years. This is the only environment that has ever produced elite operators at speed. The concepts, summarized:

• No multiple-choice tests. Students are assessed the way the military assesses: draw the circuit, run the weld, code the patch. If you can't do it, you haven't learned it.
• The Oral Board. Before advancing, students defend their work in front of instructors and peers, under pressure. Written tests can be gamed. Oral boards cannot.
• Enforced pace. We set the tempo faster than comfortable and expand the student's capacity over time. Comfort is the enemy.
• Cumulative retrieval. Week 6 quizzes demand Week 1 recall. Half of everything learned is forgotten within days without it.
• Squad System. Students move through the program in squads. Success is collective. Failure is shared. Throughout the program, squads are broken and reshuffled. A student who can only perform inside a fixed team is a student who will fail the first time a team changes.
• Stress inoculation. The first bad day happens in training, not on a live contract. By the time the pressure is real, the student has already met it.
• Zero-latency application. Teach it Tuesday morning; apply it Tuesday afternoon. Concept to usage is measured in hours, not semesters.

This is how the military has produced elite operators for a century. It is also the only framework that can produce a CNC machinist or a precision welder in months instead of years. This is how we will build the elite talent America needs.

The industrial base is not going to rebuild itself

Zane is right. This is the bet that unlocks every other bet. The trade school system America has is not the trade school system America needs.

We are building the one it needs.