Skip to content

AI Market · Forecast desk · Live · power

39%

Google accepts at least 500 MW of Fervo's conditional expansion and the parties execute a definitive agreement by June 30, 2027.

Probability history

0%25%50%75%100%09-0509-06deadline

Called at 41% on September 5, 2026. Repriced 1 time; now 39%. 297 days to deadline.

How the number was built

Final method: stacked-fixed-weights (resolved n=57 < 80). Each rung is an independent estimate; the final is a stacked combination.

Stated (issue date)41%
Reference class (n=29)50%
Recalibrated53%
Feature model35%
Council34%
Final39%

Drivers

  • Reference class qtype:disclosure runs 0.50 (n=29) vs stated 0.41.
  • Hierarchical calibration moves stated 0.41 to 0.53 (lens power, kappa_stated=24.0, kappa_lens=200.0).
  • Feature model at 0.35: conjuncts=2, named disclosure=True, horizon=298d.
  • Council of 5 at 0.34.

Cautions

  • all 5 votes came from one vendor (openai); extremization disabled, persona diversity only
  • 3 cited URL(s) failed verification and were dropped.

Council

5 personas on one model (openai/gpt-5.6-sol); other vendors were unavailable this run, so read this as one model's persona-averaged view, aggregated by log odds mean without extremization.

34%

grid and power economist · openai/gpt-5.6-sol

Google and Fervo have an established commercial relationship, so counterparty fit is stronger than for a cold-start offtake. But the rule requires acceptance—not another conditional MOU—and at least 500 MW, a very large increment. In 297 days, drilling results, financing, transmission deliverability, and contract terms must align enough for a primary disclosure. Queue COD statistics are not directly dispositive because energization is unnecessary, but long development timelines and low completion rates still reduce willingness to make a firm commitment. I therefore price this below the forecaster’s 0.41.

35%

semiconductor supply chain analyst · openai/gpt-5.6-sol

The explicit conditional expansion creates a credible path, and Google’s rising data-center power demand makes additional geothermal contracting plausible. But accepting at least 500 MW would be a major commitment relative to Fervo’s demonstrated operating scale, with drilling, financing, transmission, and interconnection risks. The 297-day window is short for converting conditional capacity into an accepted expansion. The strict requirement for acceptance—not another MOU or option—plus primary disclosure makes this a multi-leg event. I therefore place it below the forecaster’s 0.41.

36%

capital markets credit analyst · openai/gpt-5.6-sol

The 297-day window is meaningful, and Google has strategic reasons to secure firm clean power. But ≥500 MW is a large commitment to an emerging geothermal platform, and the rule requires actual acceptance—not another conditional MOU—plus primary disclosure. Such an exercise likely depends on development, financing, drilling, transmission, and pricing milestones. Interconnection data show long completion timelines, while elevated rates can complicate project financing. Contracting can precede construction, but milestone slippage and the ambiguity-against-forecaster rule put this below the relevant multi-leg and long-horizon base rates.

27%

superforecaster statistician · openai/gpt-5.6-sol

I use the 0.433 multi-leg base rate, cross-checked against 0.400 for >90-day questions and 0.500 for disclosures. I adjust downward because this requires both a large ≥500 MW acceptance and qualifying primary disclosure within 297 days; an MOU or broader partnership announcement explicitly misses. The digest provides no company-specific evidence that acceptance is imminent. Low historical queue completion adds execution caution, though acceptance can precede construction. The forecaster’s lone sub-0.50 resolution is uninformative, while broader calibration shows overconfidence, so 0.41 receives no upward deference.

37%

geopolitics regulation analyst · openai/gpt-5.6-sol

Google’s established relationship with Fervo and rising demand for firm clean power make an expansion plausible. However, accepting at least 500 MW is a large single commitment, and only 297 days remain for technical, commercial, and possibly utility or regulatory milestones. Queued-generation projects have long development timelines and high withdrawal rates, encouraging staged rather than immediate commitments. The strict resolution rule further requires an accepted expansion—not an option or conditional MOU—and a primary-party disclosure by the deadline. Those conjunctive requirements put this modestly below the multi-leg and long-horizon reference rates.

Evidence the engine used

direct · undated

News from Google | Google Product and Technology News and Stories

Cited by semiconductor-supply-chain-analyst: The explicit conditional expansion creates a credible path, and Google’s rising data-center power demand makes additional geothermal contracting plausible. But accepting at least 500 MW would be a major commitment relative to Fervo’s demonstrated operatin

direct · undated

Fervo Energy - Next-Generation Geothermal Projects

Cited by capital-markets-credit-analyst: The 297-day window is meaningful, and Google has strategic reasons to secure firm clean power. But ≥500 MW is a large commitment to an emerging geothermal platform, and the rule requires actual acceptance—not another conditional MOU—plus primary disclosure. S

direct · undated

Google Sustainability

Cited by geopolitics-regulation-analyst: Google’s established relationship with Fervo and rising demand for firm clean power make an expansion plausible. However, accepting at least 500 MW is a large single commitment, and only 297 days remain for technical, commercial, and possibly utility or regul