Pilot Zero
Pilot Zero
PROTOTYPE · CANDIDATE JUDGMENT AID · NOT VALIDATED
Primary research question
Does LBB improve users' ability to distinguish what an AI-generated representation establishes from what requires outside verification?
Candidate comparison conditions
A
Plain AI answer
The model responds directly.
B
Source / Status / Next
A simplified breakdown alongside the answer.
C
LBB
The full inspection engine — claim, status, evidence, jurisdiction, routing.
Candidate outcomes
status discriminationunsupported inferenceappropriate routingcalibrationretained understandingfresh-task transfertime burdenpreference
PREFERENCE ≠ TRUTH.
SATISFACTION ≠ ACCURACY.
Normal product use ≠ formal research participation.
Research status
Normal use of LBB is not research participation. Product feedback is not automatically enrolled in a study. Any formal research involving human participants will use a separate protocol, consent process, and any review required for that activity.
PROTOTYPE · CANDIDATE JUDGMENT AID · NOT VALIDATED
LBB has not been established to improve critical thinking, accuracy, learning, calibration, or decision-making.