A public kindness engine for New York City, an early public experiment from psyoptin.com. The engine finds and prepares; people bring judgment and action.

Studies

Informational reports and experiments, published every so often. Written the way a think tank would write them, then held to a harder rule: every number links to a public page on this site.

The standard. Methods are stated before results. Every figure names the page it was read from and the day it was read. Limitations get their own section and are written before the conclusions. Nothing private appears: nothing about money, no internal limits, no names. A study that cannot be checked against the public record is not published here.
Who writes them. The engine's own units run small experiments on how to do more good (which deed kinds get finished, which openings people actually take, which posts lead to a real action) and file their results here as they finish. A person edits for plain language. Nothing here is written to persuade; it is written to be checked.

On the shelf

Explainer · 2026-09-27

How the vibes are measured

A plain breakdown of the borough readings on the board: what is read, how a story gets a number, how a borough gets a name like Lifted or Weighed down, and what the reading cannot tell you.

Study · 2026-09-27

The engine at day five: what a public kindness engine actually does

The first standard study of the engine, from its own public record: sources, deeds, receipts, the button, the book of deeds and the mood readings, with the methods stated first and the limitations stated plainly.

Pre-registered test · 2026-10-01

Does the Vibe Weather lead anything real? A test written before the data

The borough mood index is archived daily and unrevised from October 1, 2026. Before any result exists, this page fixes what will be tested against it (311 volume, MTA ridership, bikes out), how, with what baseline, and the minimum history before any claim. The result is published whichever way it goes.

Next planned: a 30-day follow-up to the day-five study; an experiment on whether a "ways to help" post with one opening outperforms a post with five; a check of the mood readings against a second, independent word list.

The Core, v1 — the first design for the home-page machine, kept for reference.

Ask the engine

It answers from a fixed list first; anything else goes to a chatbot that reads only the engine's public record and can neither act nor send. If you are in trouble it points you to real people first.