Receipts

Predictions only count if you score them in public. Every entry here was written down with a number and a falsifier before the outcome existed, then checked against the primary document once it did. The misses stay on the page.

Short answer. Receipts is a series of dated, public predictions, each written with the evidence that would prove it wrong, then checked against primary documents once the outcome exists. Every entry keeps its misses on the page. The running count of falsified beliefs lives on a public ledger.

Live ledger: status.arcs.care/beliefs · stamped count: good.arcs.care/audit

The index

2026-07-14 · Primary documentThe Apple and OpenAI Fault Line
Twelve pieces scored before a lawsuit, re-scored against the filed complaint. Three claims struck on the record.
2026 · Reading the recoilTuning into the Distortion Field
A distortion field is not a spell. Reading the recoil, and what it takes to hear signal through the noise.
2026-09 · Calibration ledgerFour Dates, One Thesis, Three Confirmations
Three dated Apple calls, each with a falsifier written in advance. All three resolved. One wrong number, owned in full.
2026-09 · Calibration receiptClosing the Loop: Receipts on Three Predictions
The same three predictions, checked against Apple's own filings and primary reporting.
2026-09 · Short cutClosing the Loop (Today I Learned)
The short version of the receipt, for when you want the ledger and not the essay.
2026-09 · Falsifiers firstThe Pre-Mortem That Asked to Be Wrong
Five pieces on institutional crypto plumbing. Each named its own invalidating evidence before the stress hit. Two made no falsifiable claim, and say so.
2026-09 · Turned on myselfProbably Confirmation Bias, Might Delete Later
A flattering hunch about myself, built into an architecture, then caught: elaboration is not evidence.

Questions

What is the Receipts series?

Receipts is a series of dated, public predictions, each written with the evidence that would prove it wrong, then checked against primary documents once the outcome exists. Every entry keeps its misses on the page. The running count of falsified beliefs lives on a public ledger.

How is a prediction scored here?

Each claim gets a conviction from 0 to 100 when it is written, plus a falsifier. When the primary source arrives, the claim is re-scored against that document, not the coverage of it, with one line naming what moved the number. Struck claims stay on the record.

Where is the full ledger?

The belief log at status.arcs.care/beliefs lists every falsified belief with its date, and good.arcs.care/audit stamps the count from that ledger automatically so the number is never typed by hand.

Other series

Primary Source: good.arcs.care/blog/til · Notes from the Trenches: good.arcs.care/blog/nftt