Elon's Terafab will build 1 terawatt of AI compute per year, 50x current global production
Predictor: Elon Musk · ep#242 "Elon Enters the Chip Race, the S&P 500 Repricing, and Human Drivers Will Become Illegal | EP #242" · source
Prediction text
Elon's Terafab will build 1 terawatt of AI compute per year, 50x current global production | the Terafab is an objective across Tesla, XAI, SpaceX to build one terawatt of AI compute per year... Elon wants to build 50 times the current production rate of the planet.
Verbatim quote
the Terafab is an objective across Tesla, XAI, SpaceX to build one terawatt of AI compute per year... Elon wants to build 50 times the current production rate of the planet.
Predictor: Elon Musk
Calibration plot (stated vs observed)
Evidence about this node from Elon Musk 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: fab_construction_3y
Greenfield semiconductor fab construction within 3 years of announce
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
Milestone chain
- 2026-12-01 → 2027-09-30pendingTerafab Phase 1 pilot fab achieves first wafer-out at GigaTexasHow: Tesla earnings call, AI Day, or 8-K confirms first AI5/AI6 wafer produced at Austin Terafab pilot facility with verified 2nm or 14-angstrom (Intel-collab) process nodeSource: deep_research_enrichedconf 70%
- 2027-06-01 → 2028-12-31pendingVolume production ramp at Terafab reaches 100K wafer-starts/month targetHow: Tesla shareholder communications confirm sustained 100K wafer-starts/month run rate at Terafab; SemiWiki or industry trade press independently corroborateSource: deep_research_enrichedconf 45%
- 2027-01-01 → 2029-06-30pendingOptimus humanoid + Tesla vehicle compute demand validates 'edge inference' thesisHow: Tesla deploys >=1M Optimus units and >=10M FSD-equipped vehicles, each consuming dedicated AI inference silicon, justifying captive fab demand publicly cited as Terafab utilization driverSource: deep_research_enrichedconf 40%
- 2027-01-01 → 2029-12-31pendingTerafab cumulative capex disclosed exceeds $25B initial budgetHow: Tesla 10-K segment disclosures, capital expenditure footnotes, or Musk public statements reveal cumulative Terafab capex >$25B reflecting upward scope revision toward 1TW objectiveSource: deep_research_enrichedconf 55%
- 2028-01-01 → 2030-12-31pendingGlobal advanced-node capacity tracked vs Terafab share — 1 TW/yr target requires >50x current ex-TSMC capacityHow: SEMI / IC Insights aggregate report measures sub-3nm wafer-equivalent capacity additions and Terafab's contribution; falls short of 1TW/yr if industry capacity not multiplied >25x by 2030Source: deep_research_enrichedconf 55%
What if this resolves?
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
Network propagation neighbors
Top incoming (parents)
Edges that influence THIS node's belief
| Kind | Node | Their prob | P(c|s=T) | P(c|s=F) | Δ implied |
|---|---|---|---|---|---|
| prereq | CMQ_004 AGI-like models matching or outperforming human experts acro — Dario Amodei | 27.6% | 0.550 | 0.050 | -0.206 |
| prereq | CMQ_001 By 2026, AI will reach 'intern-level' capability — millions — Sam Altman | 44.8% | 0.550 | 0.050 | -0.134 |
| prereq | 248_040 Pausing AI will fail and only accelerate race dynamics. — Alex Wissner-Gross | 53.0% | 0.550 | 0.050 | -0.128 |
| killer | TK03 AI Regulatory Moratorium (EU/US Capability Freeze) | 10.0% | 0.050 | 0.550 | +0.061 |
| killer | TK02 AI Compute Supply Shock (TSMC/Taiwan Disruption) | 12.0% | 0.050 | 0.550 | +0.051 |
Top outgoing (children)
Predictions THIS node influences
| Kind | Node | Their prob | P(c|s=T) | P(c|s=F) | Δ implied |
|---|---|---|---|---|---|
| prereq | 239_009 People will be on Mars within 10 years — Elon Musk | 30.9% | 0.550 | 0.050 | -0.046 |
| prereq | 235_030 Ray Kurzweil predicts Longevity Escape Velocity (LEV) by 203 — Ray Kurzweil | 39.2% | 0.750 | 0.050 | -0.043 |
| prereq | 232_047 Mass drivers on the moon will shoot AI satellites into deep — Elon Musk | 30.0% | 0.550 | 0.050 | -0.036 |
| prereq | 239_008 Moon base will exist in 10 years — Elon Musk | 28.8% | 0.550 | 0.050 | -0.024 |
| prereq | SEM_034 True artificial general intelligence will be achieved betwee — Demis Hassabis | 28.7% | 0.550 | 0.050 | -0.023 |
Ticker exposure
Beneficiaries (14)
Adverse (7)
Prerequisites (9)
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| prereq | 232_014 | Recursive self-improvement is already here, not 12 months away. | AI | — |
| prereq | 238_009 | Recursive self-improvement is already happening now (no longer three years out) | AI | — |
| prereq | 235_038 | David Sinclair begins partial epigenetic reprogramming trials with Life Biosciences in March 2026. | Biotech/Longevity | — |
| prereq | 248_040 | Pausing AI will fail and only accelerate race dynamics. | AI | — |
| prereq | CMQ_001 | By 2026, AI will reach 'intern-level' capability — millions of virtual interns performing supervised, economically useful tasks. | AI | — |
| prereq | CMQ_004 | AGI-like models matching or outperforming human experts across most professional domains could arrive as early as 2026-2027. | AI | — |
| killer | TK01 | AGI Capability Plateau (2026-27 Training Stall) | — | — |
| killer | TK02 | AI Compute Supply Shock (TSMC/Taiwan Disruption) | — | — |
| killer | TK03 | AI Regulatory Moratorium (EU/US Capability Freeze) | — | — |
Dependents (5)
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| prereq | 235_030 | Ray Kurzweil predicts Longevity Escape Velocity (LEV) by 2033. | Biotech/Longevity | — |
| prereq | SEM_034 | True artificial general intelligence will be achieved between 2032 and 2042 — 'first we solve AI, then use AI to solve everything else'. | AI/AGI | — |
| prereq | 239_008 | Moon base will exist in 10 years | Space | — |
| prereq | 239_009 | People will be on Mars within 10 years | Space | — |
| prereq | 232_047 | Mass drivers on the moon will shoot AI satellites into deep space; self-sustaining lunar city will follow. | Space | — |
Linked documents (1)
| Sim | Source | Title | Market prob | Polarity | Reviewed | Published |
|---|---|---|---|---|---|---|
| 0.623 | arxiv | A Novel Preprocessing-Driven Approach to Remaining Useful Life (RUL) Prediction Using Temporal Convolutional Networks (TCN) | — | mentions | pending | 2026-05-04 |
Raw metadata
{
"nia": false,
"qty": "1 terawatt/year (50x global 20GW baseline)",
"url": "https://www.youtube.com/watch?v=wMLcIWLlcWg",
"mode": "CITED_PREDICTION",
"role": "Cited-Executive",
"caveats": "Elon timing predictions historically 15-20% accurate per Salim; directionally correct even if 3x longer",
"context": "the Terafab is an objective across Tesla, XAI, SpaceX to build one terawatt of AI compute per year. To put this in context, the global output today is 20 gigawatts",
"to_year": 2031,
"cited_by": "Peter Diamandis",
"verbatim": "the Terafab is an objective across Tesla, XAI, SpaceX to build one terawatt of AI compute per year... Elon wants to build 50 times the current production rate of the planet.",
"conv_cues": "audacity; objective; he's going to do it",
"direction": "NUMERIC_TARGET",
"from_year": 2028,
"timeframe": "~5 years (near-term single-digit years)",
"conv_level": "HIGH",
"milestones": [
{
"kind": "prereq",
"label": "Recursive self-improvement is already happening now (no longer three years out)",
"status": "hit",
"weight": 0.5,
"ordinal": -11,
"source_id": "238_009",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "prereq",
"label": "David Sinclair begins partial epigenetic reprogramming trials with Life Biosciences in March 2026.",
"status": "hit",
"weight": 0.5,
"ordinal": -10,
"source_id": "235_038",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "prereq",
"label": "Pausing AI will fail and only accelerate race dynamics.",
"status": "hit",
"weight": 0.5,
"ordinal": -9,
"source_id": "248_040",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "prereq",
"label": "Recursive self-improvement is already here, not 12 months away.",
"status": "hit",
"weight": 0.5,
"ordinal": -8,
"source_id": "232_014",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "prereq",
"label": "By 2026, AI will reach 'intern-level' capability — millions of virtual interns performing supervised, economically useful tasks.",
"status": "pending",
"weight": 0.5,
"ordinal": -7,
"source_id": "CMQ_001",
"expected_date": "2026-06-26",
"observed_date": null
},
{
"kind": "llm_pre_event",
"label": "Terafab Phase 1 pilot fab achieves first wafer-out at GigaTexas",
"source": "deep_research_enriched",
"status": "pending",
"weight": 0.4,
"ordinal": -6,
"source_id": null,
"confidence": 0.7,
"source_url": "https://en.wikipedia.org/wiki/Terafab",
"expected_date": "2027-05-01",
"research_origin": "deep_research",
"expected_date_range": {
"to": "2027-09-30",
"from": "2026-12-01"
},
"measurement_criterion": "Tesla earnings call, AI Day, or 8-K confirms first AI5/AI6 wafer produced at Austin Terafab pilot facility with verified 2nm or 14-angstrom (Intel-collab) process node"
},
{
"kind": "prereq",
"label": "AGI-like models matching or outperforming human experts across most professional domains could arrive as early as 2026-2027.",
"status": "pending",
"weight": 0.5,
"ordinal": -5,
"source_id": "CMQ_004",
"expected_date": "2027-07-01",
"observed_date": null
},
{
"kind": "llm_pre_event",
"label": "Volume production ramp at Terafab reaches 100K wafer-starts/month target",
"source": "deep_research_enriched",
"status": "pending",
"weight": 0.4,
"ordinal": -4,
"source_id": null,
"confidence": 0.45,
"source_url": "https://www.fintechweekly.com/news/terafab-launch-tesla-spacex-xai-chip-factory-austin-march-2026",
"expected_date": "2028-03-16"
... (truncated)