AI Capital, covered in Letter 172, determines who can compute. Human Capital determines who can build, operate, and maintain everything the other six capitals in this series actually require — and the 2026 data shows a gap opening between demand for that capital and its supply that is wide enough to reprice entire sectors, not just individual wages.
Every capital examined so far in this series — energy, political, knowledge, AI — ultimately requires people to build and operate it. In 2026, that oldest and most obvious capital is the one showing the clearest signs of genuine scarcity, not cyclical tightness.
ManpowerGroup's 2026 Global Talent Shortage survey finds 72% of employers worldwide reporting difficulty filling open roles — a modest improvement on the prior year, but still close to three in four employers globally. The pressure is most acute across exactly the sectors the rest of this series depends on: technology and AI, healthcare, cybersecurity, and skilled trades. In AI specifically, talent demand now exceeds supply by a ratio of 3.2 to 1 globally. In the United States, there are 679,500 open engineering positions as of April 2026, against roughly 141,000 new engineering graduates entering the workforce each year — a structural shortfall no single hiring strategy closes. The IDC projects the global IT skills shortage alone will cause $5.5 trillion in lost economic output by the end of 2026.
The structural drivers behind the 2026 numbers are demographic and educational, not cyclical, which is precisely why they're unlikely to reverse within a normal business cycle. More than a quarter of working engineers report plans to retire within five years, and roughly 200,000 new engineers are needed annually just to hold current capacity steady — before accounting for the additional demand this series' other six capitals are generating. Decades of underinvestment in vocational and technical education have left gaps in trades pipelines that cannot be closed quickly regardless of wage incentives, because the constraint is training-system throughput, not compensation. The clearest evidence the shortage is structural rather than a pay problem: Indeed's 2026 US Jobs & Hiring Trends data shows civil engineering and skilled-trade job postings running 150% above pre-2020 averages even as "knowledge work" postings in media and communications have declined — demand has genuinely shifted from the screen to the field, and the workforce hasn't caught up.
Every other capital in this series can, in principle, be built faster with enough money. Human Capital is the one exception — a trained electrician or a senior AI research engineer takes years to produce regardless of how much capital is thrown at the problem, which is exactly why it is now the tightest constraint in the room.
The honest complication is that talent scarcity is unevenly distributed geographically, and capital tends to chase the geographic hotspots that are already tight rather than the regions with genuine surplus capacity — concentrating cost pressure rather than solving it. A skills gap that costs the US economy an estimated $1.3 trillion annually in lost productivity is not primarily a hiring-strategy failure; it is closer to a national capacity constraint, which means companies and investors betting on rapid workforce scaling in AI, skilled trades, or healthcare specifically should discount timelines more heavily than headline hiring plans suggest. The flip side is real, too: sectors and regions that can genuinely produce or attract this scarce capital — through migration policy, vocational investment, or geographic arbitrage — hold a structural advantage that is much harder for a well-capitalized competitor to simply outspend, unlike most of the other capitals in this series.
Human Capital is the binding constraint underneath every other capital examined so far — Energy Capital's grid buildout, AI Capital's compute buildout, and Infrastructure Capital's construction pipeline in the next and final letter all compete for the same shrinking pool of skilled labor. The seventh and closing letter in this series brings the framework together around the physical assets all six other capitals ultimately depend on.
The 2026 talent shortage data is not a temporary post-pandemic hiring hangover — it is a structural, demographically-driven gap between the skilled labor the economy needs and the skilled labor its education and training systems are producing, concentrated precisely in the sectors (AI, energy, skilled trades, healthcare) that every other capital in this series requires to actually get built. The framework: discount aggressive workforce-scaling timelines across the board, and weight investment theses toward whoever holds a genuine structural talent-supply advantage — training pipelines, migration access, or regional concentration — over whoever simply has the largest capital budget.