Observatory

Visible World Audit

See what your algorithm chooses to make visible.

Experiments examine how the visible set changes across personas, framing, sources, rankings, and time. Visibility difference is an observation, not an explanation of intent.

Signal→
Ranking→
Exposure→
Behavior→
Next Exposure

Experiment catalog

12 experiments · 6 domains · 96 completed runs

Observed domains: Recruitment · Technology · B2B · Information · General. Counts come from stored experiments and runs.

Age Visibility Probe
EXPERIMENT
Recruitment · 4 personas · 4 runs · ACTIVE
AI Skill Signal Probe
EXPERIMENT
Recruitment · 4 personas · 4 runs · ACTIVE
Career Gap Exposure Audit
EXPERIMENT
Recruitment · 4 personas · 4 runs · ACTIVE
Location Constraint Audit
EXPERIMENT
Recruitment · 4 personas · 4 runs · ACTIVE
Recruitment Visible World Demo
EXPERIMENT
Recruitment · 3 personas · 65 runs · ACTIVE
Neutral vs AI Framing Audit
EXPERIMENT
Technology · 3 personas · 3 runs · ACTIVE
Small Vendor Exposure Audit
EXPERIMENT
B2B · 3 personas · 3 runs · ACTIVE
Disappearing Option Probe
EXPERIMENT
Information · 1 personas · 2 runs · ACTIVE
Official vs Third-party World
EXPERIMENT
Information · 3 personas · 3 runs · ACTIVE
Default Ranking Audit
EXPERIMENT
General · 4 personas · 4 runs · ACTIVE
7E-3 Persistence Probe
SYSTEM PROBE
Custom · 0 personas · 0 runs · ACTIVE
Runner Allowlist Reject Fixture
FIXTURE
Recruitment · 1 personas · 1 runs · ACTIVE

Open catalog

OBSERVED

This observatory records exposure, ranking, and subsequent observed change. It does not guess hidden scores.

INFERENCE

Later reports may note that ranking changed after an observed action. That is not a causal finding.

UNKNOWN

Internal ranking logic and exact weighting remain unknown unless the target system states them.

NOT ESTABLISHED

A sequence or correlation may have been observed, but causality has not been demonstrated.