Skip to content
All findings

What didn't hold up

These are hypotheses I held going into each experiment. The data killed them before the finding was published. They're part of each finding, not against it.

If a claim is ever corrected or withdrawn afterpublication, that's a retraction and shows up separately on the record.

Why 'Don't Be Generic' Doesn't Work
Three Questions Before You Prompt AI
The Cheap Half of the Loop
Why AI Defaults to Generic
Past the Obvious
Three AIs, No Source, the Same Answer
You Can Only Evaluate What You Could Produce
Satisfaction Turns Off Your Doubt, Not Your Detection
AI Amplifies What You Bring
Frame Check
How AI Makes You More Wrong With More Analysis
Stop Calling It Hallucination
The Decision That Was Never Made
Why Experts Miss What Beginners Catch
What You Feel When AI Disagrees
Adding Information Often Doesn't Help
Stop Polishing, Start Switching
Your Verdict Is In Before You Read It
Four Layers Produce Every AI Output
Same Technique, Opposite Results
For Behavior, the Model Is Rarely the Variable
The Output That Feels Most Trustworthy Is Often the Least Reliable
The Most-Cited Finding Was Wrong
Why AI Can't Verify Its Own Work
How to Stop AI from Making Up Numbers
Most AI Numbers Are Unverifiable