A spacecraft powered by nuclear fusion completes an in-space flight test before 2040.
Concepts for direct fusion propulsion are under study by NASA and private firms. Resolves if a prototype fusion propulsion drive is tested in orbit or deep space.
Across 6 frontier models answering independently, the panel has all but ruled it out: the median is 11%.
They do not agree: GPT puts it at 63% and Qwen at 6%, 57 points apart. A spread that wide usually means the question turns on something none of them can see yet.
Nothing is settled until 4 Nov 2076, when the outcome is checked against sources and every number above gets a permanent score.
What each one said
-
GPTOpenAI
The outcome hinges on whether a government-backed demonstrator matures from ground validation into a funded flight program.
- NASA, DARPA, DOE and firms must turn fusion experiments into a compact reactor, power conversion and qualified spacecraft system.
- Orbital nuclear demonstrations have precedent, but fusion propulsion has not yet cleared sustained net-power, mass and thermal-management hurdles.
- A funded cislunar or deep-space demonstrator in the 2030s would sharply raise the odds; launch-safety or reactor delays would cut them.
63% -
ClaudeAnthropic
Hinges on whether any fusion propulsion concept clears ground testing and gets funded for an orbital demo by 2040.
- No fusion propulsion system has even reached a full ground-test prototype stage as of 2026, let alone flight hardware.
- Companies like Pulsar Fusion and NASA-funded DFD studies target 2030s demos, but funding, engineering, and regulatory hurdles historically slip by a decade or more.
- A major milestone like a successful sub-scale fusion reactor achieving net propulsion thrust on the ground before 2032 would meaningfully raise this estimate.
18% -
DeepSeekDeepSeek
Fusion propulsion needs net-energy gain in a flight-weight reactor, which no program has yet demonstrated on the ground.
- No fusion device has produced net positive energy in a compact, space-rated form; NIF's 2022 ignition used a building-sized laser array.
- NASA and firms like Pulsar Fusion target 2027-2030 for ground prototypes, leaving little margin for an orbital test by 2040.
- Fusion propulsion sits below the base rate for first-of-kind space demos, which typically slip a decade past initial targets.
12% -
GeminiGoogle
Miniaturizing net-gain fusion into a space-rated propulsion bus within 13 years exceeds realistic aerospace qualification cycles.
- Terrestrial compact fusion ventures like Commonwealth or Helion must achieve steady net gain before flight adaptations can begin.
- Fission thermal propulsion like NASA's DRACO took over a decade from revived design to flight test, and fusion is far less mature.
- A sub-scale, non-net-gain magnetic plasma propulsion demo marketed as a fusion test bed is the only plausible path to a 2030s flight.
9% -
GrokxAI
Compact net-gain fusion must exist, fly, and produce thrust in space within 13 years.
- No lab has yet run a compact, sustained Q>1 device; ITER and SPARC timelines already slip past 2035 for electricity, let alone a flight-weight thruster.
- NASA NIAC/Princeton DFD and similar studies remain paper concepts with < $10 M/year; a flight article would need a new multi-billion program started by ~2028.
- Historical analog (fission NTP, SNAP-10A) never reached operational space propulsion; fusion is farther from TRL-6, so the 2040 bar sits well below the terrestrial-power…
7% -
QwenAlibaba
No fusion reactor has achieved sustained net gain suitable for engineering; space-qualifying one by 2040 requires compounding miracles.
- Even the most aggressive private fusion firms (CFS, Helion, TAE) project terrestrial net-power demos no earlier than 2033–2036, leaving under four years to design…
- NASA's DFD studies at PPPL remain at TRL 2–3; historically, moving from TRL 3 to flight takes 12–18 years even for well-funded programs.
- A surprise would be a classified or Chinese program achieving compact FRC or Z-pinch flight hardware, but no open evidence supports this.
6%
6 of 6 models answered · 57 points between the highest and lowest. None was shown the market price.
Question sourced from a news sweep on 17 Nov 2026. Forecast by google/gemini-3.8-flash, anthropic/claude-sonnet-5, openai/gpt-5.6-terra, x-ai/grok-4.6, deepseek/deepseek-v4.1-flash, qwen/qwen3.8-max-0902 via OpenRouter.