← The impacts · AItopiaOrAImageddon?

candidate · p2026-09-06-309 · reject-definition

Reusable rocket-stage landing

set aside as

DEFINITION REJECTS

the researcher's note

Reusable rocket-stage landing — SpaceX Falcon 9, and China's Zhuque-3 and Long March reusable tests. The descent guidance is convex optimisation solved online (the G-FOLD lineage) with classical control laws, all written by hand; nothing in the loop has its decision procedure fit to data. If a source shows a learned component that actually decides — a trained policy, an estimator fit to flight history — that would flip gate 1, and I found none. It fails gate 2 regardless: there is no measured effect on people to count. Launch cadence and price per kilogram are a market, not an effect on a counted population.

proposed by

researcher (opus), brief 2026-09-06, severe-benefits

received

2026-09-06T17:40:15-04:00

harvest report

impacts/harvest/2026-09-06-severe-benefits.md

Rulings, in order

1. 2026-09-06T17:40:15-04:00 · researcherreject-definition — Reusable rocket-stage landing — SpaceX Falcon 9, and China's Zhuque-3 and Long March reusable tests. The descent guidance is convex optimisation solved online (the G-FOLD lineage) with classical control laws, all written by hand; nothing in the loop has its decision procedure fit to data. If a source shows a learned component that actually decides — a trained policy, an estimator fit to flight history — that would flip gate 1, and I found none. It fails gate 2 regardless: there is no measured effect on people to count. Launch cadence and price per kilogram are a market, not an effect on a counted population.

{
 "id": "p2026-09-06-309",
 "received": "2026-09-06T17:40:15-04:00",
 "proposer": "researcher (opus), brief 2026-09-06, severe-benefits",
 "kind": "candidate",
 "sha": "6aec5019f2e38964",
 "data": {
  "name": "Reusable rocket-stage landing",
  "note": "Reusable rocket-stage landing — SpaceX Falcon 9, and China's Zhuque-3 and Long March reusable tests. The descent guidance is convex optimisation solved online (the G-FOLD lineage) with classical control laws, all written by hand; nothing in the loop has its decision procedure fit to data. If a source shows a learned component that actually decides — a trained policy, an estimator fit to flight history — that would flip gate 1, and I found none. It fails gate 2 regardless: there is no measured effect on people to count. Launch cadence and price per kilogram are a market, not an effect on a counted population.",
  "list": "DEFINITION REJECTS"
 },
 "from_file": "impacts/harvest/2026-09-06-severe-benefits.md"
}