A structured model of civilization with a prediction layer on top. Five interconnected knowledge brains tracking climate, AI, financial crisis, human behavior, and systemic power dynamics.
Perplexity answers questions. This system asks them.Perplexity and NotebookLM are reactive: you bring a question, they search and synthesize. The Great Predictor is proactive: it maintains a structured model of the world, detects when dimensions shift, traces causal chains across domains, and surfaces predictions you didn't think to ask for.
Search engines know everything about today. This system knows what connects to what, what happens next, and whether it was right.
| Capability | Perplexity | NotebookLM | Great Predictor |
|---|---|---|---|
| Knowledge base | Entire internet (shallow) | Your uploaded docs (deep but static) | 1,160+ curated entries across 5 brains (deep AND maintained) |
| Cross-domain reasoning | Only if you ask the right question | Only within uploaded sources | Automatic via 8 mapped causal chains |
| Prediction generation | User must prompt each time | Not designed for prediction | Structured prediction ledger with calibration tracking |
| Source freshness | Real-time search (no memory of what changed) | Frozen at upload time | 56 scheduled sources with refresh cycles |
| Mirage detection | None — reports headlines as-is | None — trusts source documents | Built-in per dimension (headline vs reality gap) |
| Bias flagging | Minimal | None | Active flag on every ingest |
| Calibration | None — no prediction tracking | None | Brier scores, Tetlock methodology |
| Persistence | New session = blank slate | Persists per notebook | Cumulative across sessions, brains cross-reference |
Each brain has a distinct prediction domain. They don't all predict the same way.
Each brain's prediction power is different, and they chain together. The most consequential predictions are cross-domain — no single expert connects Australian droughts to Arab Spring to EU climate policy collapse. The causal chain architecture does this structurally.
31 tipping points tracked across four domains, each with current state vs threshold, proximity, reversibility, and cascade connections.
| Tipping Point | Domain | Status |
|---|---|---|
| Commercial Real Estate Vacancy | Financial | AT THRESHOLD (20% vs 20-25% trigger) |
| Institutional Trust | Social | AT THRESHOLD (US 30-35% vs <30% trigger) |
| Disinformation Detection | Technology | CROSSED (2024-2025) |
| AI Job Automation (copywriting/coding) | Technology | CROSSED (2022-2024) |
| Coral Reef Die-off | Climate | CROSSING (0-10 years) |
| Permafrost Abrupt Thaw | Climate | REGIONAL TRIGGERS MET (0-15 years) |
| Tipping Point | Domain | Proximity |
|---|---|---|
| Property Insurance Withdrawal | Financial | 2-5 years to systemic mortgage impact |
| Private Credit / Shadow Banking | Financial | 2-4 years |
| AI Compute vs Grid Capacity | Technology | 2-4 years globally |
| Centola Climate Minority (25%) | Social | 3-5 years |
| Affective Polarization Limit | Social | 1-3 election cycles |
| AI Job Automation (customer service) | Technology | Crossing now |
| Breadth | Perplexity searches the entire internet. We have 1,160 curated entries — deep but narrow to our five domains. |
| Speed to first answer | Perplexity answers in seconds. This system requires brain-building first. |
| Zero setup | Both competitors work out of the box. This system required building brains, mapping dimensions, scheduling sources. |
| Breaking news | Perplexity is real-time. Our source refresh is scheduled, not live. |
| General knowledge | Anything outside our five domains, Perplexity wins entirely. |
The investment compounds. Every brain entry, every resolved prediction, every source refresh, every mirage detection makes the system smarter. Perplexity's value resets every session. NotebookLM's value is bounded by what you uploaded.
After six months of operation:
| Perplexity | Knows exactly as much as it did on day one about your specific domains. |
| NotebookLM | Knows what you manually fed it, with no cross-referencing. |
| Great Predictor | 1,500+ curated entries, 50+ tracked predictions with calibration data, refined mirage checks, proven causal chains, and a behavioral model tested against real events. |
The engine is an appreciating asset. The others are utilities.
Perplexity is a search engine with AI. NotebookLM is a document reader with AI. The Great Predictor is a mental model of civilization with AI. They're playing different games.
The question isn't "which is better." It's "what are you trying to do?" If you want answers, use Perplexity. If you want to understand your documents, use NotebookLM. If you want to predict what happens next and know whether you're right — this is the only option that even attempts it.