ai-for-less-suffering

Public thinking on AI deployment toward suffering reduction. A typed reasoning tool that separates descriptive convergence from normative divergence, and finds coalition bridges that do not require conceding the divergence.

Node types

P

Camp --- Coherent cluster of held claims.

CD

Descriptive claim --- What is.

CN

Normative claim --- What should be.

S

Source --- A citable primary document.

W

Warrant --- A source's stance on a claim --- support, contradict, or qualify.

I

Intervention --- A proposed action.

LF

Friction layer --- What slows deployment.

LH

Harm layer --- A welfare harm to weigh.

Worked example

What the site does, shown against a real claim currently in the graph. The same descriptive statement carries warrants that support, contradict, and qualify it --- so the disagreement is visible as evidence, not as vibes. Click any source or claim to traverse.

The graph

• Analyses 7

  Palantir coalition, leverage rank, steelman of a target.

• Camps 8

  Coherent clusters of held claims and divergent anchors.

• Claims 43

  30 descriptive · 13 normative.

• Interventions 4

  Proposed actions scored against friction and harm layers.

• Sources 42

  Citable primary docs, datasets, filings, and press.

• Layers 11

  5 friction · 6 harm.