In case you missed it ...
☀️ ✨ The strange divorce of nuclear fusion and super AI wagers (Monday)
🦅🐼 Sizing up US-China AI strategy: A Quick Q&A with … Kyle Chan (Wednesday)
⤵️ The degrowth United Nations (Thursday)
☀️ ✨ The strange divorce of nuclear fusion and super AI wagers (September 29, 2025)
Plasma Promise. Nuclear fusion was once the future of energy. After decades of missed deadlines, it’s finally seeing real momentum — just as prediction markets remain far more bullish on artificial intelligence than on fusion’s near-term prospects. Which is weird.
Flashback. In the 1960s, it seemed like the miracle of fusion technology was just around the corner. Sci-fi writer Arthur C. Clarke predicted fusion would solve the world’s energy needs by 2000. The RAND Corporation thought it would arrive by the 1980s. Alas, the years passed with no sign of viable fusion energy, relegating the tech to the cynical conclusion: It was “always a few decades away.”
Breakthrough. In 2022, Lawrence Livermore scientists achieved the first net energy gain from fusion. That accomplishment has proven replicable, leaving private firms to take up the challenge of scaling.
Some major contendors:
Commonwealth Fusion Systems is about 60 percent finished with its SPARC device. They raised $800m in 2025, and has billion-dollar power deals with Google and Italy’s oil major, Eni.
Helion Energy is being backed by Sam Altman. They broke ground on a Washington state plant in July, promising power for Microsoft by 2028.
Zap Energy has logged over three hours of continuous plasma operation, validated by DOE.
Outlook. Despite the influx of funding, prediction markets remain skeptical. Metaculus forecasters placed the odds of achieving AGI by 2030 at around 50 percent, but expect the first 100 MW fusion plant sometime around 2043. Manifold traders were very similar: AGI soon, fusion much later. But shouldn’t we assume that powerful AI will actually help us make major progress towards a fusion reality? There could be a few reasons why this gets lost:
Fusion’s many false promises of years past have generated skepticism.
AGI contracts hinge on benchmarks; fusion contracts demand real-world grid contributions and a tangle of regulators and financiers.
AI grabs headlines and consumers’ attention; fusion hides in lab reports and engineering schematics.
Even if AGI arrives, scaling fusion to reliable, grid-ready power will be a slow, brutal grind.
Up Wing Up Shot. If AGI or superintelligence really does arrive by the 2030s, it should speed fusion’s timeline, despite markets’ failure to factor that in. If AI-fusion mutual reinforcement is real, fusion energy might actually be just “a few decades away” — or sooner.
🦅🐼 Sizing up US-China AI strategy: A Quick Q&A with … Kyle Chan (October 1, 2025)
Keep reading with a 7-day free trial
Subscribe to Faster, Please! to keep reading this post and get 7 days of free access to the full post archives.