Every bias in Part 2 follows this exact format. Here is Bias #1 so you can see the depth and structure.
Bias #1: Reward and Punishment Super-Response
Definition
We overweight immediate incentives and punishments, distorting evaluation of long-term outcomes.
The Trap
You optimize for the metric that gives fast feedback instead of the outcome that matters most.
AI Manifestation
You accept the first AI output that "looks right" because fixing it feels like punishment, even when you know it needs another pass.
Root Cause (Klesha)
Raga — attachment to pleasure. The desire for the quick win overrides your critical eye.
Shortcut
"Before I accept this, what would I change if I had to iterate once more?"
Litmus Test
If removing the time pressure would change your evaluation, the bias is active.