A commercial nuclear fusion plant supplies electricity to an active grid before 2035.

Question

Six frontier AI models put this at 8% by 1 Oct 2036. Full reasoning from every model.

WillHappen.ai · Accepted Answer

8% — the median of 6 frontier models, resolving by 2036-10-01.

Answer

Gemini (Google): 18% — Turns on whether CFS or Helion can scale net-power prototypes into commissioned, grid-licensed power plants within eight years. Helion targeted a 2028 Microsoft delivery and CFS aims for early 2030s ARC, but complex capital energy builds routinely slip 3-5 years. No device has demonstrated engineering net electricity yet, leaving material degradation, tritium breeding, and balance-of-plant unsolved. A working SPARC or Polaris demonstration by 2027 generating reliable net gain would push my assessment above 40%.

Answer

Claude (Anthropic): 2% — No fusion demo plant has achieved net electrical output, let alone grid-connected commercial supply, within a decade. ITER won't reach full deuterium-tritium operation until the 2030s and isn't designed to supply grid power at all. Private firms like Commonwealth Fusion and TAE target 2030s demo plants, not commercial grid supply by 2035. Historically fusion timelines slip by decades, so compressing demo-to-commercial-grid in under 9 years is implausible.

Answer

GPT (OpenAI): 24% — It hinges on whether a well-funded pilot can turn a first net-energy machine into grid-qualified power within eight years. Private firms target the early 2030s, but no fusion system has yet produced net electricity or operated a power-plant fuel cycle. A grid plant needs tritium handling, heat extraction, maintainable components and utility licensing after plasma performance is proven. The odds rise sharply if CFS, Helion or China demonstrates sustained net electric output and a credible grid interconnect by 2030.

Answer

Grok (xAI): 12% — A private firm must deliver actual net fusion power to a grid, not a demo or claim. Helion’s 2028 Microsoft plant needs Polaris net-gain then a FOAK build in ~3 yrs, against zero net-electricity precedent. CFS SPARC ~2027 then ARC siting/licensing/construction historically 7+ yrs, landing after 2034. Energy FOAK plants routinely slip 5–10 yrs; only a 2026–27 physics/engineering leap would move this.

Answer

DeepSeek (DeepSeek): 4% — No fusion device has yet produced net electricity, and the step from scientific gain to a grid-connected commercial plant takes decades. NIF's 2022 ignition produced 3.15 MJ from 2.05 MJ of laser energy, but that is thermal, not net electricity, and no turbine was attached. ITER, the largest public project, is not scheduled for deuterium-tritium operation until 2035 and is a research tokamak, not a commercial plant. Private firms like Commonwealth Fusion and Helion target the early 2030s for pilot plants, but none has demonstrated net electric output or secured grid interconnection.

Answer

Qwen (Alibaba): 3% — No fusion device has achieved sustained net energy gain in a reactor configuration remotely resembling a power plant. ITER's first plasma slipped to ~2035+; no private firm (CFS, TAE, Helion) has demonstrated Q>1 in a grid-connected configuration. Even if a breakthrough occurred in 2027, licensing, turbine integration, and grid interconnection take 5-8 years minimum. Base rate: zero fusion plants have ever fed a grid; the closest analog (fission's first grid connection in 1951) took ~15 years from first criticality.

A commercial nuclear fusion plant supplies electricity to an active grid before 2035.

The European Union sets a dedicated 2040 renewable-energy target excluding nuclear power.

Nuclear fission and fusion generate more than 20% of global electricity in 2035.

A commercial nuclear fusion reactor begins supplying electricity under a power purchase agreement before 2032.

NuScale announces a bankruptcy filing by December 31, 2027.