Skip to content

AI Market · Forecast desk · Live · power

52%

A power service agreement for the Paducah AI campus is filed with the Kentucky Public Service Commission by December 31, 2026.

Probability history

0%25%50%75%100%08-0109-06deadline

Called at 63% on August 1, 2026. Repriced 1 time; now 52%. 116 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)63%
Reference class (n=29)50%
Recalibrated55%
Feature model64%
Council45%
Final52%

Drivers

  • Reference class qtype:disclosure runs 0.50 (n=29) vs stated 0.63.
  • Hierarchical calibration moves stated 0.63 to 0.55 (lens power, kappa_stated=24.0, kappa_lens=200.0).
  • Feature model at 0.64: conjuncts=1, named disclosure=True, horizon=152d.
  • Council of 5 at 0.45.

Cautions

  • all 5 votes came from one vendor (openai); extremization disabled, persona diversity only

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.

43%

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

The exact-document rule is a meaningful hurdle: negotiation or announcement does not count without a file-stamped PSA or equivalent. No docket number, executed agreement, or published filing schedule is identified. The question also does not establish that the serving utility must submit the campus agreement to the PSC rather than handle it through another approval path. Four months is enough for an advanced special-contract filing, but large-load agreements typically require load, security, cost-allocation, and infrastructure terms first. The forecaster’s >90-day and disclosure base rates are only 0.4 and 0.5, and its 0.60–0.69 forecasts have realized below stated confidence.

48%

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

A year-end filing remains plausible, but the resolution requires a file-stamped PSA or equivalent in a specific Kentucky PSC docket, not merely a utility announcement or signed private agreement. With only 116 days left, negotiation, load-study, tariff, and commission-filing preparation create meaningful schedule risk. The power reference class is favorable but very small; the broader >90-day and disclosure classes are near or below 0.5. The forecaster’s 0.60–0.69 band has also realized below stated confidence. No Paducah-specific filing milestone is supplied, and ambiguity resolves against a hit.

46%

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

A filing requires an executed or filing-ready agreement, utility approval processes, and a docket stamp within 116 days; general AI-power demand does not establish those project-specific steps. Paducah-area service may involve municipal or TVA-related arrangements, so Kentucky PSC filing jurisdiction is not automatic. The deadline is still achievable if negotiations are already advanced, because the rule requires filing rather than approval or energization. I weight the >90-day and disclosure base rates near 0.4–0.5 more heavily than the small power sample, putting this below the forecaster’s 0.63.

41%

superforecaster statistician · openai/gpt-5.6-sol

I use a 0.45 base rate, weighting the >90-day class (0.40) and disclosure class (0.50); the 0.70 power rate has only four observations. The forecaster’s 0.60–0.69 band realized 0.556, indicating overconfidence relative to 0.63. This requires a particular regulatory filing within 116 days, so agreement timing alone is insufficient, and ambiguity over an “equivalent” resolves against a hit. The digest provides no project-specific, verifiable filing milestone warranting an upward adjustment.

48%

geopolitics regulation analyst · openai/gpt-5.6-sol

The filing requires two milestones within 116 days: execution of a qualifying service arrangement and a Kentucky PSC file stamp clearly tied to the Paducah campus. Special-contract filings can follow execution quickly, so construction or interconnection completion is unnecessary. However, the supplied evidence contains no project-specific docket, executed agreement, or announced filing schedule. The forecaster’s comparable disclosure and >90-day cases resolved near 0.5 and 0.4, respectively; the stronger power-lens record has only four observations. Year-end timing and the rule resolving unclear equivalents or campus attribution against the forecast further reduce the chance of a hit.

Evidence the engine used

direct · undated

KY PSC Home

Cited by grid-and-power-economist: The exact-document rule is a meaningful hurdle: negotiation or announcement does not count without a file-stamped PSA or equivalent. No docket number, executed agreement, or published filing schedule is identified. The question also does not establish that the serv