The 2030s Convergence
A conditional forecast that renewable energy and affordable local AI/robotics may mature together in the 2030s.
Forecast & Model · Initial edition, November 2026
The forecast is that within roughly the next decade, two great technological transitions may become materially mature at once: renewable electricity with storage, and inexpensive intelligence that can operate locally, eventually including general-purpose personal robotics.
This is a proposed convergence, not a quantitative forecast yet.
Starting conditions
In the mid-2020s, energy systems remain mixed. Solar, wind and batteries are increasingly important but do not yet replace fossil fuels across the full energy economy. Frontier AI is highly capable but costly and concentrated in data centres. Useful local models exist, but their capability and operating economics remain open questions. General-purpose robotics remains far from routine, reliable, inexpensive deployment.
Meanwhile, the neoliberal political settlement and US-led international order appear increasingly unstable.
The proposed convergence
A future in which much energy can be generated close to use, much everyday intelligence can be run under the user's control, and some flexible physical work can be performed by affordable robots might challenge the economics of large central infrastructures and distant production networks.
The forecast is not that local AI becomes smarter than the frontier. It may merely become good enough for an expanding fraction of ordinary tasks. The cloud frontier could continue racing away—or plateau—without making the local-sufficiency hypothesis irrelevant.
Why dates are uncertain
Model quality cannot be reduced to one benchmark. Local deployment depends on memory, bandwidth, energy consumption, accelerator costs and software efficiency. Robotics adds reliability and manipulation challenges that may not track language-model progress.
Renewable electricity shares and batteries also do not translate immediately into complete electrification of industry, transport and heating.
What we need to measure
Track capability of locally runnable models on real tasks, hardware cost and performance per watt, useful robotics autonomy, solar-plus-storage economics, grid and transport electrification, and the share of global trade whose costs could be avoided by local production.
A range of outcomes is possible. The 2030s Rupture is the stronger hypothesis about the consequences if several thresholds are crossed close together. This forecast should be revised against actual evidence rather than defended as a fixed deadline.