Continued expansion of fossil-fuel extraction infrastructure
degradingNew oil, gas, and coal extraction projects (licensing, pipelines, LNG terminals, refineries) on a scale consistent with current trajectories — the infrastructure commitments made by major producers and importers between roughly 2024 and 2030. The configuration-change is the lock-in of physical and financial capital with multi-decade amortisation horizons.
A biosphere already operating near several planetary boundaries (binding-constraint: atmospheric and oceanic sinks are the close walls, not solar influx). R_living for most major classes of life is bounded by climate stability within roughly the Holocene envelope, on which functional diversity and current ecosystem geometry depend.
Additional cumulative emissions and longer use-lives for the marginal infrastructure. Probability mass shifts toward warming trajectories above the thresholds at which coral, low-altitude polar, montane, and major-river-delta ecosystems lose their current configurations. R_living for those classes contracts toward zero on multi-decade horizons; R_living for most other classes narrows by removing the lower-warming branches of the reachable set.
Centuries to millennia for the climatic effects (CO₂ residence time, ice-sheet hysteresis). Decades for the infrastructure-financial lock-in.
Most major classes of life, with steepest loss to ecosystems already at the edge of their climatic envelope. Human communities in low-lying coastal and arid-margin regions disproportionately. Future generations who inherit a configuration with permanently narrower R_living (time-asymmetry).
Existing fossil-capital holders, in financial terms, over the lock-in window. Some short-horizon energy-security observer-classes. Nothing at category level for life-persisting observers (observer-relative-option-space).
A demonstrated, scaled atmospheric-sink technology with no net-additional planetary-boundary cost — a substitution for the absorption-side invariant flagged in substitution-limits. Not promises; demonstrated. Or evidence that current emissions trajectories do not in fact narrow R_living for the named ecosystem classes — the IPCC's own confidence intervals would have to shift substantially. Neither is currently in evidence.
The textbook case for an unambiguous Δω-negative verdict. The move adds configurations to the reachable set (more energy throughput at the financial-board score) while subtracting them at every longer-horizon, larger-class-of-life board it is also being played on. The viable-objective no-regret form rejects moves that asymptotically empty R_living for some major class at some relevant horizon; this move does that for several. The verdict survives observer-relativity: even adopting human-only observer-classes, long-horizon humans lose more than short-horizon humans gain. The rule produces the conventional environmental verdict here — but it produces it by the framework's own logic, not by importing one.