The West’s workforce is shrinking and its tax system still leans almost entirely on salaries. On this week’s Bricks, Bucks & Bytes rundown, Foundamental’s Patric Hellermann proposed a two-part fix: capitalize automation deployments aggressively enough to push Western industrial automation spend from $300 billion toward $1 trillion a year, then shift taxation off labor and onto the robots doing the work. Martin pushed back on sequencing, and the disagreement is where this piece gets useful.
Every automation conversation eventually runs into the same wall: who pays for society when the workers retire? The US construction industry alone was short about 439,000 workers in late 2025, per the Information Technology and Innovation Foundation, and Associated Builders and Contractors sees the need climbing toward half a million extra workers by 2027. The people replacing them, at current pipeline rates, do not exist.
On this week’s episode, Patric Hellermann connected that demographic squeeze to a policy idea that sounds provocative until you hear the logic. Then it sounds provocative and hard to dismiss.
The Tax Base Is Retiring
Labor carries the load, and labor is walking out the door
Martin set up the problem with a reference to a David Freeberg interview he’d recently heard: in most Western economies, capital is taxed lightly and labor is taxed heavily. Payroll, income, national insurance, social contributions. The whole fiscal machine assumes a large, working, salaried population.
Demographics are quietly breaking that assumption. As workers retire faster than they’re replaced, the tax base leaves with them. You can raise rates on the shrinking pool, which accelerates the squeeze, or you can find something else to tax. Patric’s answer is sitting on the factory floor.
Tax the robots, guys. It’s not hard. It’s really not hard. Just tax the robots and build them like absolutely crazy. And you’re going to flourish and people can actually retire at 55.Patric Hellermann, General Partner, Foundamental
Dustin took the same idea from the political angle. If voters want expanded public programs, someone has to fund them, and a heavily automated, robot-taxed industrial base is a mechanism that actually scales. Whatever you think of the politics, the arithmetic is coherent: productivity has to come from somewhere, and it increasingly won’t come from headcount.
Capitalize the Deployments First
The $1 trillion path runs through contractors, not humanoids
The tax shift is step two. Step one, in Patric’s telling, is getting the robots installed at all. He estimates the West currently spends around $300 billion a year on industrial automation, roughly 75% of it deployed through projects: configuration, simulation, programming, wiring, commissioning. That work is done by contractors and integrators, and they are undercapitalized for the volume the demographic math demands.
His prescription: “If you capitalize deployments super aggressively, you can go from 300 billion to a trillion annual spend and actually really take all of our Western factories into a high state of automation.” Financing, insurance, and servicing for the firms doing the installs, rather than yet another funding round for hardware that Asia already builds better and cheaper.
The pattern shows up in construction robotics too. The machines winning real deployments are narrow, practical tools with clear ROI, as we covered in our look at where robots are actually working on sites in 2026. The constraint is rarely the robot. It’s whether anyone can deploy, support, and finance it at scale.
Workforce, robots, and the money behind both
One weekly briefing on the forces deciding who builds what, and with whom.
Join 3000+ ReadersMartin’s Counterargument: Sequence Matters
Requalification money spent before the robots work is money burned
Martin didn’t dispute the destination. He disputed the order of operations. His view: first you need workable solutions actually showing up on sites and in factories. Only then does it make sense to rebuild education and requalification systems around them. Spend heavily on retraining now, before the automation is proven, and you’re preparing people for jobs that don’t exist yet in workflows nobody has settled. His estimate for turning the education system around once the tech lands: probably a decade.
It’s a fair challenge, and there’s a version of the answer already running. The companies most exposed to the labor gap aren’t waiting for policy. Meta committed $115 million to a free skilled-trades academy with a job guarantee, and Amazon runs its own training pipeline for data center construction roles – we broke down why big tech is suddenly paying to train electricians in July. The market is patching the near-term gap with human training while the automation bet matures. Patric’s proposal is about what happens after both curves cross.
Was anyone fully convinced by the end of the segment? Not really. But the disagreement maps the actual policy choice better than most think tank papers manage.
| Lever | Flagship example | Open question |
|---|---|---|
| Train more humans now | Meta’s $115M Workforce Academy with pre-training job offers | Can pipelines scale before the retirement wave peaks? |
| Automate and shift the tax | Hellermann’s capitalize-deployments-then-tax-robots proposal | Do robots work at scale before the tax base erodes? |
Two coordinated moves. First, aggressively capitalize automation deployments – financing and servicing for the contractors and integrators who install industrial automation – to grow Western spend from roughly $300 billion to $1 trillion a year. Second, for policymakers: reduce taxes on salaries and tax the robots instead, so government revenue doesn’t shrink as the workforce retires. He argues the combination could let people retire at 55.
The US industry was roughly 439,000 workers short as of late 2025, according to the Information Technology and Innovation Foundation, with the gap concentrated in electricians, pipe layers, and HVAC trades. Associated Builders and Contractors projects the need rising toward nearly half a million additional workers by 2027. (Source)
That’s the standard objection, and it’s why sequencing dominates the debate. Hellermann’s version front-loads the incentive: pour capital into deployments first so automation reaches scale, then apply the tax once robots are productive enough to carry it. Critics would argue any robot tax raises the hurdle rate for exactly the investment he wants. The episode didn’t settle it, and neither has anyone else.
Increasingly, the project owners themselves. Meta launched a $115 million skilled-trades academy with free training and conditional job offers before day one, and Amazon runs training programs tied to its data center construction pipeline. The owners funding the labor pipeline is a genuinely new pattern in construction. (Source)
Related Articles
Why Big Tech Is Suddenly Paying to Train Electricians
Construction Robotics in 2026: 4 Workflows Where Robots Are Actually Working
All3 Bags $25m to Roboticise Everything Above the Foundation
