Manufacturing’s 2030 decision is a causal chain, not a 2×2.
The standard scenario framework treats geopolitical fragmentation and autonomy as independent axes. They aren’t. Fragmentation is causing the autonomy push. Collapse the four scenarios to two trajectories with one switching condition.
The dominant analyst framework for manufacturing’s 2030 future is a 2×2: geopolitical fragmentation versus global integration on one axis, autonomous versus incremental automation on the other. Four scenarios pop out — squeeze-out factories, human-led factories, globalized factory ecosystem, regional closed dark factories. Product leaders are told to design for portability across all four. The framework is clear, the visual is satisfying, and the underlying logic is wrong: the two axes are not independent. Fragmentation is what is causing the autonomy push. Treating them as orthogonal hides the actual decision the operator has to make.
This piece argues a structural reframe. Collapse the 2×2 into a causal chain. The destination is determined by where each region’s manufacturing economics land in the labor-availability and policy-regime equation, and the autonomy decision is downstream of that. Two trajectories matter, not four. Designing for “portability across all four scenarios” is a McKinsey-grade hedge that lets analysts off the hook and leaves operators paying for optionality they will not use.
Why the axes aren’t independent.
The autonomy push in manufacturing is not happening in a vacuum. It is happening because three structural forces have aligned: labor scarcity in OECD economies (US labor force flat by 2029, Japan and Germany already shrinking), capital intensity that only works at “lights-out” or near-lights-out scale, and regional policy regimes that make supply-chain portability the new competitive value. Fragmentation drives all three. When the US imposes a 25% tariff on Chinese components, manufacturing reshores to North America. Reshored manufacturing in a labor-constrained economy can only pencil at autonomy levels significantly higher than the legacy global supply chain assumed. The autonomy is not optional; it is the only way the unit economics work in the new policy regime.
Read this way, the 2×2 collapses. There is no plausible 2030 scenario where the world has fragmented and manufacturers have not pushed harder on autonomy — fragmentation forecloses the human-led and incremental-automation quadrants. There is also no plausible 2030 scenario where the world has reglobalized and manufacturers have leaned hard on autonomy in OECD geographies — reglobalization shifts capacity back to lower-labor-cost regions where autonomy economics are weaker. Two of the four 2×2 quadrants are not on the table. The remaining two are the actual trajectories.
Two trajectories, one switch.
Trajectory A: Fragmentation persists. Reshored OECD capacity goes high-autonomy.
Tariffs hold, geopolitical risk premiums on cross-border supply chains stay elevated, and reshoring continues. OECD manufacturers invest heavily in autonomous capacity because the labor math forces the issue and the unit economics demand high utilization. The squeeze-out factories and regional dark factories of the conventional 2×2 collapse into one trajectory: regional, autonomous, capital-intensive. Autonomy is not a choice; it is the only way the new geography of production pencils.
Trajectory B: Re-globalization. Capacity shifts back to low-labor-cost regions, autonomy investment softens.
Tariffs ease, geopolitical risk declines, capital flows return to lower-labor-cost manufacturing geographies. The autonomy investment thesis weakens because the labor arbitrage returns. The globalized factory ecosystem and human-led quadrants of the conventional 2×2 collapse into one trajectory: distributed, lower-autonomy, labor-leveraged. This is the soft case for autonomy and the strong case for incremental automation in OECD geographies — autonomous deployment becomes optional, not mandatory.
The switching condition is geopolitical, not technological.
Which trajectory plays out by 2030 is determined almost entirely by trade policy and security posture, not by AI capability. AI capability is the same in both trajectories; what differs is whether the unit economics demand its deployment in OECD geographies. Operators planning capital deployment cannot hedge across both trajectories — they can only place a probabilistic bet on which trade regime persists. Pretending the four quadrants are equally likely is a hedge with a real cost: the operator overinvests in optionality and underinvests in the trajectory that actually arrives.
The strongest argument against this position.
The strongest counter is that trade regimes can shift unpredictably, and the operator should hedge by designing for portability. This is the conventional wisdom and it has structural appeal — flexibility seems risk-reducing. But portability hedging has a real cost that is rarely calculated: every architectural decision designed for “works in all four quadrants” is more expensive, slower to ship, and lower-performing than a decision designed for the most likely single trajectory. The operator who hedges loses to the operator who places the bet correctly. Portability is the right hedge for an operator who genuinely cannot read the trade regime; it is the wrong hedge for an operator who can read it and is using portability as an excuse not to commit. Most current portability investment is the second case.
Three things to do this quarter.
01 · Pick a trajectory. Build a 5-year capital plan against the trajectory you actually believe in, not against a 2×2 hedge. If you believe fragmentation persists, plan for autonomous OECD reshoring at scale. If you believe reglobalization, plan for distributed lower-autonomy capacity. The hedge is a decision you have not yet made.
02 · Identify the switching signal you will watch. The signal is not AI capability or vendor maturity. The signal is the trade regime — tariff levels, content rules, security posture, capital flow restrictions. Name the three indicators you will watch quarterly, and pre-commit to what you will do if the signal flips.
03 · Stop paying for portability you will not use. Most “portable architecture” spend in manufacturing today is hedge against scenarios that are not on the table. Audit your portability investments against the two-trajectory framework, and reallocate the spend that is hedging against unrealistic quadrants into the trajectory you have actually picked.
SALT’s position-review rhythm grades published positions against subsequent reality.