Observed arrival · 2026-10-11
John Henry puts human judgment against AI agents
A pilot compares work by an expert using AI with work by an autonomous AI agent, then asks reviewers to judge anonymous outputs.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Experts evaluating AI-assisted work and invited reviewers
- Worth noticing
- Reviewers set criteria before work begins, then receive anonymous outputs in randomized order.
Field notes
The protocol separates setting the standard from judging the work: reviewers define useful output before the task, and both sides receive the resulting criteria. The autonomous agent is allowed to plan, research, and revise within the same declared spending limit as the expert’s work. The visible healthcare-strategy example is labeled as an interface demonstration, not evidence from a completed comparison.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue