A free Monte Carlo prediction engine
The Kraken.
The Newsroom Prediction Engine.
Drop a dataset, organization, event, or trend. The Kraken runs 50 narrative framings through five interpretive lenses, clusters them into 10 stable attractors, and predicts the stories most likely to emerge from a real newsroom.
5 interpretive lenses · 50 parallel framings · 10 ranked attractors · 6-section report
How the Kraken works
- Coverage research. The Kraken first pulls recent press coverage and dataset references about your input. This grounds the predictions in what has actually been written.
- Five lens passes. Five parallel calls — each carrying a different interpretive lens (economic, cultural, structural, behavioral, technological) — generate 10 distinct framings apiece. 50 framings total.
- Clustering. The 50 framings are clustered into the 10 strongest narrative attractors — the stable gravity wells that emerge from the simulation.
- Editorial Success Value. Each cluster is scored on clarity, novelty against saturation, emotional activation, cross-audience appeal, diffusion potential, rejection resistance, and evidence groundability.
- Six-section report. The top clusters are expanded into a complete report: cluster map, top 10 stories, top 10 pitches, top 10 headlines, top 10 insights, and top 10 publication targets.
Total runtime: ~45 seconds. Total cost to 701am: about three cents.
The Kraken is free. The cost is ours. The output is yours.
The Kraken is waking
This usually takes about 45 seconds
★ The Kraken returned empty-handed ★
Something went wrong.
An unknown error occurred.
★ The Kraken Report ★
Filed at sunrise · Saturday · 16 May 2026
—
A predictive map of the stories most likely to emerge from this input, ranked by editorial success value.
§ 01 · Cluster map
Ten stable attractors.
—
§ 02 · Top 10 stories
The narratives that survive.
Ten finalized stories. Each scored against editorial filters. Each ready to refine.
§ 03 · Top 10 pitches
Ten pitches, publication-ready.
Each pitch is structured for the journalist's inbox — subject line, three sentences, no fluff.
§ 04 · Top 10 headline sets
Headlines, by format.
Five variations per story, formatted for different outlet styles.
§ 05 · Top 10 non-obvious insights
What the data quietly reveals.
Signals that survive editorial scrutiny and still feel discovered, not deduced.
§ 06 · Top 10 publication targets
Where the work will most likely land.
For each story, the outlet archetype most likely to publish it — and the framing adjustment that wins acceptance.