← Cockpit
CMQ_017predictionEnergypower-grid

US electricity production must grow by 'tens of percent' by end of decade to support hundreds of millions of frontier-AI GPUs.

Predictor: Leopold Aschenbrenner

Prior probability
75.0%
Current probability
51.9%
evolves via intake + LBP
Conviction
5/5
Signal quality
A
Resolution
in_progress
Window
2026-01-01 – 2026-11-30
Edges in / out
4 / 0
Tickers exposed
22

Prediction text

US electricity production must grow by 'tens of percent' by end of decade to support hundreds of millions of frontier-AI GPUs. | tens of percent | US electricity production growth rate

Key catalyst: US electricity production growth rate

Watch events: US electricity production YoY; SMR deployment; behind-the-meter gas turbine deployments at AI campuses.

Verbatim quote

From episode "The Global Architecture of Machine Intelligence: Exhaustive Synthesis of AI Compute, Memory & Quantum Predictions (2023-2026)"
tens of percent

Resolution evidence

Status: in_progress

Hyperscaler 5GW+ site contracts across PA, TX, OK, AZ; Microsoft Three Mile Island nuclear restart; Meta nuclear RFP 2024; AWS/Google SMR deals 2025-2026.

Predictor: Leopold Aschenbrenner

κ + Brier as of 2026-05-22
κ (discount)
0.688
Brier
0.0417
excellent
Hits / Misses
2 / 0
of 3 resolved
Hit rate
66.7%
Calibration plot (stated vs observed)

Evidence about this node from Leopold Aschenbrenner is multiplied by κ in /api/intake. Lower κ = less weight; floors at 0.10 (effectively silenced) and caps at 1.00 (full weight).

Reference class: energy_grid_rebuild_5y

Linked

Major grid expansion (>50GW) within 5y of announce

Base rate
50.0%
3/6 historical
Inside weight
0.615
TRF=0.55
Outside weight
0.385
pulling toward base rate
inside 53.0% → blend 51.9% -1.2pp)

Tetlock-style outside view: at TRF=1 (just predicted), outside view dominates (w_in=0.3). At TRF=0 (deadline), inside view dominates (w_in=1.0). The blend regularizes overconfident inside views toward the historical base rate.

Probability over time

7 prob_history rows
0%25%50%75%100%prior 75%2026-05-022026-05-172026-05-30
intake v2milestone miss sweeplbp propagationreference class assignedlegacy v1prior_prob (analyst seed)current = 51.9%

Milestone chain

Pre-event signals (upstream prereqs + window checkpoints) → resolution event → downstream cascades. Status/dates update from linked nodes; re-derive nightly via scripts/ops/derive_milestones.py.
Leading chain: 2 overdue ⏱ · 1 pending
  1. 2026-03-09overdueQ1 window check-in (25%)
  2. 2026-05-16overdueQ2 window check-in (50%)
  3. 2026-07-23pendingQ3 window check-in (75%)

No downstream cascades — this prediction is a leaf in the dependency graph.

What if this resolves?

Clamp this prediction TRUE or FALSE and run a counterfactual Gibbs sample. Surfaces the predictions whose marginals shift most under that assumption.
(live posterior: 52%)

Click a button to clamp this prediction and run a Gibbs sample. Returns the predictions whose marginals shift most. ~30s per run; ideal for stress-testing "if X resolves, what else moves?"

Evidence chain

Every probability update with full Bayesian provenance — chronological, latest first
metadata_milestone_miss_sweep2026-05-30T22:15:00Z51.9%-8.0pp
metadata_milestone_miss_sweep bayesian_v2 n=1 inside=0.530 blend=0.519 LLR=-0.279 κ=0.69 w_in=0.62 energy_grid_rebuild_5y
Raw metadata
{
  "trf": 0.5497684922277138,
  "kappa": 0.6875,
  "base_rate": 0.5,
  "predictor": "Leopold Aschenbrenner",
  "total_llr": -0.4054651081081644,
  "grace_days": 7,
  "bayesian_v2": true,
  "prior_logit": 0.4006779028515317,
  "bayes_factor": "1.3:1 against",
  "blend_reason": "blend 61% inside / 38% outside (TRF=0.550, base_rate=0.500 from energy_grid_rebuild_5y)",
  "inside_prior": 0.598850522655255,
  "kappa_source": "predictor_table",
  "n_milestones": 1,
  "blend_applied": true,
  "contributions": [
    {
      "llr": -0.4054651081081644,
      "kind": "quartile_checkpoint",
      "kappa": 0.6875,
      "label": "Q2 window check-in (50%)",
      "weight": 0.05,
      "strength": "weak",
      "confidence": null,
      "source_url": null,
      "adjusted_llr": -0.278757261824363,
      "expected_date": "2026-05-16",
      "measurement_criterion": null
    }
  ],
  "evidence_kind": "metadata_milestone_miss_sweep",
  "inside_source": "history_v2",
  "inside_weight": 0.6151620554406003,
  "outside_weight": 0.3848379445593997,
  "posterior_prob": 0.5187414535778369,
  "posterior_logit": 0.12192064102716871,
  "predictor_brier": 0.04167,
  "inside_posterior": 0.5304424599050337,
  "blended_posterior": 0.5187414535778369,
  "reference_class_id": "energy_grid_rebuild_5y",
  "total_adjusted_llr": -0.278757261824363,
  "predictor_n_resolved": 3
}
LBP2026-05-24T02:00:02Z59.9%-13.0pp
Network propagation: 72.8% → 59.9%
4-iter LBP, residual 0.01000 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run 806b02f8
intake_event_update2026-05-21T23:15:16Z72.8%+25.8pp
intake:7afeeb9a-f217-4dd2-b910-24ff14bdfc39 bayesian_v2 inside=0.728 blend=0.728 LLR=1.106 κ=0.69 no_blend
Raw metadata
{
  "trf": 0.5766698637873985,
  "kappa": 0.6875,
  "base_rate": null,
  "predictor": "Leopold Aschenbrenner",
  "total_llr": 1.6094379124341,
  "bayesian_v2": true,
  "prior_logit": -0.1196129791451904,
  "bayes_factor": "3.0:1 favoring",
  "blend_reason": "no reference_class linked",
  "inside_prior": 0.4701323570814465,
  "kappa_source": "predictor_table",
  "blend_applied": false,
  "contributions": [
    {
      "llr": 1.6094379124341,
      "kappa": 0.6875,
      "label": "ERCOT 10%/yr growth and PJM 76% YoY wholesale price spike confirm 'tens of percent' electricity demand expansion thesis.",
      "adjusted_llr": 1.106488564798444
    }
  ],
  "evidence_kind": "intake_event_update",
  "inside_source": "history_v2",
  "inside_weight": 1,
  "outside_weight": 0,
  "posterior_prob": 0.7284703504805086,
  "evidence_origin": "daily_intake",
  "llm_suggestions": [
    {
      "polarity": "corroborates",
      "status_change": "unchanged",
      "evidence_strength": "strong",
      "delta_prob_suggestion": 0.07
    }
  ],
  "posterior_logit": 0.9868755856532536,
  "predictor_brier": 0.04167,
  "evidence_doc_ids": [],
  "inside_posterior": 0.7284703504805086,
  "blended_posterior": 0.7284703504805086,
  "reference_class_id": null,
  "total_adjusted_llr": 1.106488564798444,
  "predictor_n_resolved": 3
}
LBP2026-05-17T02:00:01Z47.0%-3.4pp
Network propagation: 50.4% → 47.0%
5-iter LBP, residual 0.00689 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run e607fa96
LBP2026-05-10T02:00:02Z50.4%-6.7pp
Network propagation: 57.1% → 50.4%
6-iter LBP, residual 0.00584 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run e5c18d29
LBP2026-05-03T02:00:01Z57.1%-12.3pp
Network propagation: 69.4% → 57.1%
6-iter LBP, residual 0.00677 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run 1a683ac9
metadata_milestone_miss_sweep2026-05-02T22:07:21Z69.4%-5.6pp
metadata_milestone_miss_sweep bayesian_v2 n=1 inside=0.694 blend=0.694 LLR=-0.279 κ=0.69 no_blend
Raw metadata
{
  "trf": 0.6338685427014515,
  "kappa": 0.6875,
  "base_rate": null,
  "predictor": "Leopold Aschenbrenner",
  "total_llr": -0.4054651081081644,
  "grace_days": 7,
  "bayesian_v2": true,
  "prior_logit": 1.0986122886681098,
  "bayes_factor": "1.3:1 against",
  "blend_reason": "no reference_class linked",
  "inside_prior": 0.75,
  "kappa_source": "predictor_table",
  "n_milestones": 1,
  "blend_applied": false,
  "contributions": [
    {
      "llr": -0.4054651081081644,
      "kind": "quartile_checkpoint",
      "kappa": 0.6875,
      "label": "Q1 window check-in (25%)",
      "weight": 0.05,
      "strength": "weak",
      "confidence": null,
      "source_url": null,
      "adjusted_llr": -0.278757261824363,
      "expected_date": "2026-03-09",
      "measurement_criterion": null
    }
  ],
  "evidence_kind": "metadata_milestone_miss_sweep",
  "inside_source": "prior_prob",
  "inside_weight": 0.5562920201089838,
  "outside_weight": 0.4437079798910162,
  "posterior_prob": 0.6942055654642756,
  "posterior_logit": 0.8198550268437468,
  "predictor_brier": 0.04167,
  "inside_posterior": 0.6942055654642756,
  "blended_posterior": 0.6942055654642756,
  "reference_class_id": null,
  "total_adjusted_llr": -0.278757261824363,
  "predictor_n_resolved": 3
}

Network propagation neighbors

Top edges sorted by latest LBP cross-impact
All propagation →

No propagation data yet. Run inference/.venv/bin/python scripts/ops/run_loopy_belief_propagation.py on the droplet, or wait for the Sunday 02:00 UTC weekly cron.

Ticker exposure

22 ticker(s) linked

Beneficiaries (13)

VRTFLNCFSLRARGANSITMETNGEVHTHIYSBGSYHUBBPWRCMISMNEY

Adverse (3)

ARCHBTUCEIX

Prerequisites (4)

Predictions that must hit first
TypePredTitleDomainLag
correlateS_GRID_50GW_202750GW dedicated AI/data center grid by Dec 2027energy_grid_expansion
correlateS_COMPUTE_10GW_2028Compute: 10GW total by Dec 2028compute_scale
correlateS_COMPUTE_100GW_2030Compute: 100GW national-scale by Dec 2030compute_scale
correlateS_GRID_50GW_DELAYED50GW grid expansion delayed past 2031energy_grid_expansion

Dependents (0)

Predictions enabled by this
TypePredTitleDomainLag
No dependents

Expected milestones (1)

From Sheet 17 Monitoring Triggers
Expected byDescriptionStatus
2026-12-31[Energy/Regulation 2026-12] rscaler PPA announcements; BYOP permits [CMQ_017] US electricity production YoY; SMR deployment; behind-the-meter gas turbine deployments at AI campus [235_022] US will add a record 86 gigawatts of utility-scale capacity this coming year.pending

Validations (1)

Resolution events
Observed atStatusByNotes
2026-04-29partialthesis_timeline_v1.0_importHyperscaler 5GW+ site contracts across PA, TX, OK, AZ; Microsoft Three Mile Island nuclear restart; Meta nuclear RFP 2024; AWS/Google SMR deals 2025-2026.

Raw metadata

From Thesis_Timeline_v1.0_FINAL workbook
{
  "nia": false,
  "qty": "+tens of % US elec production",
  "mode": "FORECAST",
  "role": "Cited-Researcher",
  "caveats": "Assumes continued training + inference compute expansion at projected rates.",
  "context": "Identifies power as the single most inflexible physical constraint on AGI compute buildout.",
  "to_year": 2030,
  "verbatim": "tens of percent",
  "conv_cues": "must grow; explicit magnitude",
  "direction": "UP",
  "from_year": 2026,
  "timeframe": "by 2030",
  "conv_level": "HIGH",
  "milestones": [
    {
      "kind": "quartile_checkpoint",
      "label": "Q1 window check-in (25%)",
      "status": "overdue",
      "weight": 0.05,
      "ordinal": -3,
      "source_id": null,
      "expected_date": "2026-03-09",
      "observed_date": null,
      "miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
      "miss_emitted_by": "metadata_milestone_sweep"
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q2 window check-in (50%)",
      "status": "overdue",
      "weight": 0.05,
      "ordinal": -2,
      "source_id": null,
      "expected_date": "2026-05-16",
      "observed_date": null,
      "miss_emitted_at": "2026-05-30T22:15:00.756418+00:00",
      "miss_emitted_by": "metadata_milestone_sweep"
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q3 window check-in (75%)",
      "status": "pending",
      "weight": 0.05,
      "ordinal": -1,
      "source_id": null,
      "expected_date": "2026-07-23",
      "observed_date": null
    },
    {
      "kind": "event",
      "label": "US electricity production must grow by 'tens of percent' by end of decade to support hundreds of millions of frontier-AI GPUs.",
      "status": "pending",
      "weight": 1,
      "ordinal": 0,
      "source_id": "CMQ_017",
      "expected_date": "2026-09-29",
      "observed_date": null
    }
  ],
  "repeat_eps": 1,
  "affiliation": "Situational Awareness LP",
  "attribution": "FIRST_PERSON",
  "granularity": "YEAR",
  "source_refs": "10",
  "target_date": "2029-12-15T00:00:00",
  "display_date": "2026-09-29",
  "episode_date": "2026-04-21T00:00:00",
  "key_catalyst": "US electricity production growth rate",
  "parse_method": "Report midpoint",
  "domain_bucket": "Energy",
  "episode_title": "The Global Architecture of Machine Intelligence: Exhaustive Synthesis of AI Compute, Memory & Quantum Predictions (2023-2026)",
  "flag_repeated": false,
  "in_5yr_window": true,
  "source_report": "AI_Chip__Compute__Memory__Quantum_Predictions.md (2026-04-21)",
  "appears_in_eps": "CMQ-RPT",
  "futurist_phase": "Phase 2 (2027-2028)",
  "is_macro_claim": false,
  "total_mentions": 1,
  "priority_weight": 5,
  "ps_cluster_tags": [
    "C3",
    "C4"
  ],
  "report_evidence": "Power = THE gating variable for AGI buildout; Aschenbrenner is right to flag it as primary bottleneck over chips.",
  "active_end_month": "2026-12",
  "recent_statement": "Aschenbrenner 2025-2026 commentary continually emphasizes power as binding constraint ahead of compute.",
  "watch_events_raw": "US electricity production YoY; SMR deployment; behind-the-meter gas turbine deployments at AI campuses.",
  "months_from_today": 44,
  "probability_layer": "Higher (in-flight)",
  "active_start_month": "2026-01",
  "december_dispersal": {
    "reason": "december_dispersal: domain=Energy → 11/2026",
    "new_date": "2026-11-30",
    "old_date": "2026-12-31",
    "applied_at": "2026-04-30T16:28:34.304992+00:00"
  },
  "flag_nia_bracketed": false,
  "track_record_grade": "A",
  "track_record_notes": "Physical-grid claim is one of most-testable in Situational Awareness framework; tracking confirms binding constraint.",
  "contradicting_notes": "US historical electricity production was flat 2007-2022; 'tens of percent' in 4-5 years requires wartime-pace permitting.",
  "flag_near_term_2027": false,
  "flag_high_conviction": true,
  "milestones_phase2_at": "2026-05-01T18:30:21.232773+00:00",
  "milestones_derived_at": "2026-05-02T03:08:50.593691+00:00"
... (truncated)