A timeline planning app that combines:
- Base forecast timeline (AI Futures Model + Future Timeline style predictions)
- Live signal feed from X API and Polymarket API
- Probability overlay engine that shifts month/domain probabilities from incoming market + event signals
- New Signals tab shows significant events in near-real-time
- Includes source, domain mapping, significance score, target month, and delta impact
- Displays domain-level rollup of net pressure (
+/-percentage points)
- Live signals now affect the timeline probabilities via bounded deltas
- Effects decay over subsequent months (default 3-month horizon)
- Signal items are also injected into month predictions as
[Signal] ...entries with source links
- New script:
scripts/update-live-signals.mjs - Pulls:
- X API (
/2/tweets/search/recent) whenX_BEARER_TOKENis provided - Polymarket Gamma API (
/markets) for active markets
- X API (
- Applies a strict post-AGI implication relevance filter:
- keeps signals tied to AI capability/compute/governance/energy/supply-chain implications
- removes unrelated noise (sports/celebrity/general non-implication markets)
- Writes snapshot to:
public/data/live-signals.json
npm install
npm run signals:update
npm run devOpen the app and go to the Signals tab.
Create a .env file (optional but recommended):
# Required for X ingestion
X_BEARER_TOKEN=...
# Optional tuning
X_QUERY=(AGI OR ASI OR "artificial general intelligence" OR OpenAI OR Anthropic OR DeepMind OR xAI) lang:en -is:retweet
X_MAX_RESULTS=40
POLYMARKET_API_BASE=https://gamma-api.polymarket.com
POLYMARKET_LIMIT=400
SIGNAL_WINDOW_HOURS=24
SIGNAL_MAX_ITEMS=120If X_BEARER_TOKEN is missing, the script still runs and ingests Polymarket-only signals.
npm run dev
npm run build
npm run signals:update
npm run signals:update:with-build- This is a decision-support signal layer, not an oracle.
- Deltas are intentionally bounded to avoid runaway probability swings.
- Improve calibration by tuning keyword maps, significance formulas, and per-domain weighting over time.