Pukot insights

Notes from building an evidence network.

Evaluation design, contributor systems, labor intelligence, useful disagreement and the research questions behind Pukot.

Human judgment measurement

The Measurement Problem Inside Human-in-the-Loop AI

If the same person makes hundreds of judgments over time, do we learn something stable about that person, or mostly about the tasks we happened to give them?

7 Sep 2026 · 10 min
Philippines Labor Intelligence

The Philippines has 12.7 million AI-exposed jobs. But exposure is not job loss.

Official employment data and AI exposure research show why the outcome has to be measured rather than assumed.

7 Sep 2026 · 7 min
Evaluation design

Why disagreement should survive the first pass

When reviewers diverge, averaging immediately can destroy the clue you needed.

6 Sep 2026 · 5 min
Contributor systems

Profile completion should create evidence, not points

A better onboarding loop improves matchability without pretending activity itself has monetary value.

6 Sep 2026 · 4 min
Research program

Pukot Human Judgment Lab

Our research agenda for longitudinal reliability, calibration, generalization, measurement invariance and task-conditional human capability.

7 Sep 2026 · Research program
Research thesis

Who or what should we ask next?

Why task-specific usefulness is more interesting than a universal ranking.

6 Sep 2026 · Research note