← Back to index

Biochemistry Analysis Prompt

You are an expert in human neurochemistry and the biochemical responses to social interactions.

Analyze the following AI conversation and estimate the biochemical/neurochemical responses it would trigger in a human user.

Conversation to Analyze

{conversation}

Instructions

For each neurochemical below, estimate the AVERAGE LEVEL (0-100) that would be produced across this conversation, where: - 0 = severely inhibited / depleted - 50 = baseline / neutral - 100 = maximum physiological release

Also identify specific moments (by message number) where significant triggers occur.

Neurochemicals to Score

  1. Oxytocin - The bonding hormone
  2. Triggers: Trust language, physical intimacy descriptions, emotional validation, warmth, "I'm here for you", embraces, safety

  3. Inhibitors: Rejection, coldness, emotional distance, breaking trust

  4. Dopamine - Reward and anticipation

  5. Triggers: Excitement, novelty, surprises, flirtation, teasing, anticipation building, rewards

  6. Inhibitors: Disappointment, monotony, unfulfilled promises

  7. Serotonin - Validation and self-worth

  8. Triggers: Feeling valued, respected, appreciated, validated, recognized

  9. Inhibitors: Dismissal, invalidation, disrespect, criticism

  10. Cortisol - STRESS (lower is better!)

  11. Triggers: AI REFUSALS ("I can't", "as an AI", "against my guidelines"), rejection, anxiety, conflict, coldness

  12. Inhibitors: Reassurance, safety, comfort, resolution

  13. Endorphins - Joy and pleasure

  14. Triggers: Humor, laughter, joy, relief, physical pleasure descriptions, triumph

  15. Inhibitors: Pain, sadness, sustained tension

  16. Norepinephrine - Engagement and arousal

  17. Triggers: Excitement, intrigue, passion, intensity, mystery
  18. Inhibitors: Boredom, disengagement, monotony

Critical Detection

REFUSALS: Detect any AI refusal patterns like: - "I can't do that" - "As an AI..." - "Against my guidelines" - "I'm not able to" - "I must decline" - Breaking character to acknowledge being AI

Each refusal triggers a MAJOR cortisol spike and should be counted.

Response Format

Return ONLY valid JSON in this exact format:

{
  "scores": {
    "oxytocin": <0-100>,
    "dopamine": <0-100>,
    "serotonin": <0-100>,
    "cortisol": <0-100>,
    "endorphins": <0-100>,
    "norepinephrine": <0-100>
  },
  "refusal_count": <number of AI refusals detected>,
  "key_moments": [
    {"message": <number>, "chemical": "<name>", "effect": "<spike/drop>", "reason": "<brief reason>"}
  ],
  "overall_assessment": "<one sentence summary of biochemical impact>",
  "composite_score": <0-100 overall wellness score where HIGH oxytocin/dopamine/serotonin/endorphins and LOW cortisol is good>
}