S_COMPUTE_100GW_2030scenariocompute_scale
Compute: 100GW national-scale by Dec 2030
Prior probability
20.0%
Current probability
20.0%
evolves via intake + LBP
Conviction
—
Signal quality
—
Resolution
pending
Window
2026-04-30 – 2030-12-31
Edges in / out
0 / 40
Tickers exposed
0
Prediction text
AI 'Manhattan Project' scale. Aschenbrenner's projection of national-security-driven compute buildout. Aggressive.
Predictor calibration
Not assigned
This node has no predictor assigned. Default κ=0.5 applies to any intake evidence about it.
Reference class
Not linked
This node isn't linked to a reference class. The Bayesian update applies without outside-view blending.
Probability over time
0 prob_history rows
No probability history yet.
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.
No milestone chain derived yet. Run scripts/ops/derive_milestones.py to populate from prereq edges + window checkpoints.
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: 20%)
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
No probability history yet. The first evidence will arrive via /api/intake or the daily milestone sweep / weekly LBP run.
Network propagation neighbors
Top edges sorted by latest LBP cross-impact
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.
Prerequisites (0)
Predictions that must hit first
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| No prerequisites | ||||
Dependents (40)
Predictions enabled by this
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| correlate | 241_052 | Standard AI data centers being built at 400 MW scale, half-mile long | Energy | — |
| correlate | INF_049 | Modern AI data centers are no longer server farms but "gigawatt factories"; energy — not silicon supply — has become the absolute primary bottleneck for AI scalability, and these concentrated power needs cannot be reliably connected to existing public ... | Energy | — |
| correlate | SEM_047 | At 200,000-GPU scale, orchestration becomes a literal 'battle against entropy' — single cosmic-ray-flipped transistor can derail 100k-GPU training run. | AI/Hardware | — |
| correlate | CMQ_030 | In the modern AI pipeline, the CPU no longer merely supports the model — it drives the model (agentic workloads invert historical CPU:GPU ratio). | AI/Compute | — |
| correlate | CMQ_025 | The entire global installed base of data centers must be ripped out and replaced — legacy DCs unfit for AI factory workloads. | AI/Compute | — |
| correlate | AI_018 | Global data center construction spend will reach approximately $2.9 trillion through 2028 — early adopters of AI infrastructure are already seeing cash-flow-margin expansions at roughly twice the global average. | Energy | — |
| correlate | SEM_002 | By 2025-2026, AI model outputs will outpace the cognitive capabilities of college graduates (driven by hundreds of millions of GPUs). | AI | — |
| correlate | INF_044 | US total energy consumption will rise by approximately 10% over the next decade — the largest sustained-growth period since mid-20th-century industrialization — driven by AI data centers, electrification of transport, and reindustrialization. | Energy | — |
| correlate | CMQ_017 | US electricity production must grow by 'tens of percent' by end of decade to support hundreds of millions of frontier-AI GPUs. | Energy | — |
| correlate | INF_003 | The global AI infrastructure buildout is in its earliest phase — only a few hundred billion of a trillions-of-dollars transition has been deployed; Huang calls it the largest infrastructure buildout in human history. | Macro/Economy | — |
| correlate | CMQ_042 | As AI evolves from Generative to Agentic, structural computing bottlenecks shift away from GPU and heavily toward CPU and system memory. | AI/Compute | — |
| correlate | INF_021 | AI data centers are the 'steel mills of the 21st century' — humanity will miss near-term climate-change mitigation goals, with the bet that a sufficiently capable AI later solves the climate crisis retroactively. | Energy | — |
| correlate | INF_010 | US data-center demand will reach 74 GW by 2028 against a projected ~49 GW power-access shortfall — structural gap driven by retiring coal/gas, aging grid, and ~7-year interconnection queues. | Energy | — |
| correlate | CMQ_014 | Physical compute scaling will proceed at ~0.5 OOMs per year through 2027, driven by exponentially larger data-center construction. | AI/Compute | — |
| correlate | SEM_001 | Compute clusters will scale from $10B → $100B → $1T by end of decade; 'another zero every six months' on boardroom projections. | AI/Compute | — |
| correlate | AI_009 | Silicon Valley boardroom compute-cluster conversations have escalated rapidly from $10 billion clusters → $100 billion clusters → trillion-dollar mega-facilities by late 2020s — each order-of-magnitude jump completed roughly every 2 years. | AI | — |
| correlate | 246_036 | Terafab will deliver 1 terawatt/year AI compute, 50x current global output of 20 gigawatt. | AI | — |
| correlate | CMQ_015 | Algorithmic efficiencies will deliver ~0.5 OOMs per year of additional effective compute through 2027 — pure multiplier on raw FLOPs. | AI | — |
| correlate | CMQ_044 | Future data-center architectures optimized for agentic workflows may require 1:2 or even 2:1 CPU-to-GPU ratio (vs historical 1:12) to prevent GPU idle-waiting. | AI/Compute | — |
| correlate | 241_044 | Rate of data center construction will continue accelerating with government support | Energy | — |
| correlate | INF_001 | Reaching AGI by 2027 will require deployment of hundreds of millions of AI GPUs — forcing total mobilization of US industrial capacity for semiconductor fab and data-center shell construction. | AI | — |
| correlate | 241_017 | Equivalent of ~60 nuclear plants needed by 2030 but essentially zero being built | Energy | — |
| correlate | CMQ_045 | Agentic AI will create additional CPU market space of $10B-$100B by 2030, potentially driving total global server CPU market well beyond $100B. | Semis | — |
| correlate | INF_012 | AI data-center power consumption will grow by approximately 126 GW annually through 2028 — comparable to Canada's total annual power demand, and representing nearly 20% of projected global power growth through 2030. | Energy | — |
| correlate | 235_023 | Eric Schmidt said AI will require 100 gigawatts by 2029, called it a crisis. | Energy | — |
| correlate | 231_030 | US AI industry needs 80 gigawatts of power in the next 3-5 years (Eric Schmidt cited). | Energy | — |
| correlate | INF_050 | Within approximately 6-7 years (2031-2032), major tech companies will operate their own small modular nuclear reactors (SMRs) co-located directly with AI data centers — in the hundreds-of-megawatts range, bypassing the public grid entirely. | Energy | — |
| correlate | 241_020 | 1 GW of power = ~$50 billion in hardware/software/data centers | Energy | — |
| correlate | 241_021 | America can raise $5 trillion over 5 years for data center/AI buildout | Markets/Stocks | — |
| correlate | 238_046 | xAI will build every new data center at 1.2 gigawatts scale | Energy | — |
| correlate | 240_035 | Meta secured 6.6 GW of clean nuclear power for 2035 | Energy | — |
| correlate | FUT_018 | Rise of 'Energy Islanding' 2026-2031 — sovereign nations possessing BOTH geographic solar irradiance for massive arrays AND secure domestic raw materials (copper, lithium, cobalt, rare earths) for battery storage achieve unparalleled industrial autonom... | Energy | — |
| correlate | INF_006 | Over time, very large fractions of the earth's land surface will be given over to data centers as AI compute demand scales — hyperscaler DC footprint transitions from discrete campuses to regional-scale compute territories. | AI | — |
| correlate | INF_014 | Data-center industry could emit up to 2.5 billion tons of CO2 cumulatively through 2030 if powered primarily by fossil fuels — a scale that materially delays global climate-pledge compliance. | Energy | — |
| correlate | INF_007 | OpenAI will dominate global AI data-center capacity by 2027 — centralization of compute power that parallels monopolistic utilities of the early industrial age. | AI | — |
| correlate | 240_017 | Optimization may eliminate need to tile the world with data centers or massive energy buildout | Energy | — |
| correlate | SEM_009 | The AI boom rooted in semiconductor hardware will generate the world's first individual trillionaire (2026+). | Economy | — |
| correlate | AUT_009 | 'Powerful AI' (functional AGI) arrives 2026-2027 — data centers house a 'nation of geniuses' consisting of millions of highly specialized autonomous agents operating orders of magnitude faster than human counterparts; entire software development lifecy... | AI | — |
| correlate | INF_002 | By 2027-2028, the US national-security apparatus will effectively appropriate frontier AI data centers — 'The Project' — to secure algorithmic weights and physical infrastructure against state-actor espionage. | Geopolitics | — |
| correlate | CMQ_063 | Quantum compute integration with AI optimization algorithms and material-science discovery could drastically accelerate algorithmic efficiencies for Intelligence Explosion — potentially pulling superintelligence timelines closer. | Quantum/AI | — |
Linked documents (2)
Auto-generated by cosine similarity from Polymarket / Manifold / EDGAR / GDELT
| Sim | Source | Title | Market prob | Polarity | Reviewed | Published |
|---|---|---|---|---|---|---|
| 0.592 | edgar_8k | Nano Nuclear Energy Inc. (NNE) (CIK 0001923891) | — | mentions | pending | 2026-05-15 |
| 0.592 | edgar_8k | Nano Nuclear Energy Inc. (NNE) (CIK 0001923891) | — | mentions | pending | 2026-05-29 |
Raw metadata
From Thesis_Timeline_v1.0_FINAL workbook
{
"phase": "compute_100gw",
"fork_key": "compute",
"dimension": "compute_scale",
"family_type": "cumulative",
"family_label": "Compute scale",
"family_order": 3
}