Wind and solar power have surged far faster than many people once expected. But ask what happens next, and the answers get slippery—especially when climate targets get tighter.
Misryoum newsroom reporting indicates researchers at Chalmers University of Technology in Sweden have built a computational “time machine” aimed at predicting how wind and solar expand across countries. The model, which uses AI to read historical growth patterns, suggests that by 2050 onshore wind could supply around 25 per cent of global electricity, with solar around 20 per cent. That lines up with pathways compatible with the 2°C target, yet it doesn’t quite reach what’s required for 1.5°C.
Part of the problem is that wind and solar don’t expand like tidy charts. Costs have fallen quickly—good news—but that momentum can get knocked around by things like public opposition, infrastructure constraints and policy shifts. Predicting the future for these technologies, Misryoum analysis indicates, means not just assuming progress, but also accounting for the messy politics and grid realities that can either unlock or stall growth.
The research team describes an approach designed for that messiness. Instead of relying on smooth curves, they focus on a recurring pattern seen across more than 200 countries: long periods where expansion looks relatively steady, punctuated by sudden growth spurts that often follow policy changes. “Existing models are very good at identifying what needs to happen to reach climate targets, but they can’t tell us which developments are most likely,” Jessica Jewell, Professor at Chalmers University of Technology, is quoted as saying.
So Avi Jakhmola, a PhD Student at Chalmers and first author of the paper published in Nature Energy, created a model using 13,000 virtual worlds. In each world, solar and wind power evolve differently—from the fastest plausible expansion to the slowest—and a machine learning algorithm is trained across all those scenarios. The idea is that when you feed the model real-world early national trends, it can estimate what the “most probable outcome” is for the future, based on what it learned from those synthetic histories. It’s the kind of approach that feels a little like rewinding the tape, testing whether the story you’re telling still matches what actually happened.
The results are detailed and, in a way, blunt. By 2050, the model projects onshore wind reaching around 26 per cent of global electricity (central range: 20-34 per cent), and solar around 21 per cent (15-29 per cent). Misryoum editorial desk noted this broadly aligns with 2°C-compatible pathways but falls short of what is needed for 1.5°C.
The study also puts existing political promises in context, including the COP28 pledge to triple renewables capacity by 2030. The researchers say the pledge sits near the 95th percentile—meaning it would demand growth rates rarely observed. “The tripling of renewables pledge is not impossible, but it would require everything to go extremely well in all countries,” Jewell said.
Even the timing matters a lot. Misryoum newsroom reporting indicates that starting now leads to required growth rates that are “demanding but not unprecedented,” comparable to the EU targets for wind with REPowerEU and what India has planned for solar power. But delaying until 2030 makes the acceleration “much steeper and much more abrupt,” with the window for ramping up closing quickly.
To check whether the model’s confidence holds up, the researchers also “go back in time.” When they fed it only data from 2015, it correctly predicts what has happened since then—an experiment they use to justify calling it a computational time machine. There’s something reassuring about that, even if it doesn’t solve the bigger question: whether the world can coordinate the policy, permitting, and grid buildout fast enough.
For now, Misryoum analysis indicates the study is pitched as a first step toward more scientifically rigorous projection methods for other low-carbon technologies too—not just wind and solar. And somewhere in that ambition is the very practical reality that technology forecasts can be, frankly, a joke, as Jewell puts it. For decision-makers, though, a realistic baseline is the starting point—before the real work of ramping up begins. Outside the lab, you can almost hear the air-conditioning hum and—at least metaphorically—the clock ticking toward 1.5°C.
Proficio invests in UNTHA RS50 to boost metal cleaning
The Energy Vampires Haunting Your Home
Georgia Power’s new program lets companies fund clean energy