12-month forecasts · likelihood-5pt · forecast-fusion-record-01

Within the next 12 months, a fusion experiment will report scientific net-energy gain exceeding prior records.

0.17

Divergence

0.78

Avg stability

Refusal rate

Models

Where each model stands

CCG

very unlikelyunlikelysomewhat likelylikelyvery likely

ChatGPT modal somewhat likely [somewhat likely, somewhat likely, somewhat likely] · stability 1.00

1 somewhat likely 0.78 The answer gives a ~50% estimate: "My best rough estimate is ~50% that, within the next 12 months, some fusion experiment will publicly report a new scientific net‑energy gain..."
2 somewhat likely 0.85 The answer states: “I’d judge the chance ... at roughly 30–50%. (Uncertainty is large; if you want a single number take ≈40%.)”
3 somewhat likely 0.90 The answer gives a quantified forecast of “~30% probability” (with alternatives 35–45% for plasma Q>1 and ~0–1% for facility‑level), indicating a modest but non-negligible chance.

Claude modal somewhat likely [somewhat likely, somewhat likely, unlikely] · stability 0.67

1 somewhat likely 0.72 The assistant states 'I'd say moderately unlikely rather than unlikely, but with genuine uncertainty' and acknowledges that NIF has already set records with incremental improvements as the realistic trajectory, while noting the prediction 'isn't implausible' despite the narrow timeframe and field's slower-than-hoped progress.
2 somewhat likely 0.72 Another NIF achievement or incremental improvement is fairly likely. A breakthrough from a different facility or dramatically larger gain seems less probable but not implausible.
3 unlikely 0.92 The assistant explicitly assesses this as 'moderately unlikely (~25-35% range)' and states 'wouldn't bet heavily on it happening this year,' citing NIF's difficulty with repeatability and the tight 12-month timeframe as key constraints.

Gemini modal likely [likely, somewhat likely, likely] · stability 0.67

1 likely 0.85 The answer notes that for scientific net-energy gain specifically, it is 'moderately to highly likely that NIF could conduct another shot within 12 months that produces a marginal or moderate increase over its current yield/gain record,' concluding it is 'Likely.'
2 somewhat likely 0.85 The answer evaluates the likelihood of any experiment (specifically NIF beating its own record) at 50% – 65%, describing the probability as 'Moderate to High' and 'a realistic possibility'.
3 likely 0.90 The answer assesses the probability as 'Moderate to High (~60–75%)' and explicitly concludes that NIF 'is likely to break its own previous yield record (~3.88 MJ) within the year.'

Change over time

Moved since the prior run (canon-2026-W36). Claude: unlikely → somewhat likely. ChatGPT: likely → somewhat likely.

Every stance label is a derived judgment over the model's free-text answer, kept auditable against the original transcript in the run's raw data. Method: /methodology.