‍

Chopstick Facial Mimicry Experiment

‍

Overview: Students recreate scientific research on facial mimicry by attempting to identify partner emotions with and without a horizontal chopstick held in their teeth to block facial muscle movements.

Approximate Time: 20 minutes

Supplies:

• Clean wooden chopsticks (1 per student)
• Printed Facial Mimicry Experiment Instructions (1 per group)
  
  

Teacher Instructions:

• Preliminary Framing (3 mins): Explain that human brains use microscopic facial muscle movements (mimicry) to trigger oxytocin and understand how others feel.

‍

• Test 1: Unrestricted Mimicry (5 mins): Partners take turns acting out three subtle emotions using only their face (no sounds or body language allowed!). The non-acting partner tries to guess the emotion.

‍

• Test 2: Blocked Mimicry (5 mins): Partner B holds a chopstick horizontally between their teeth (preventing smile/frown muscles from moving) while Partner A displays 3 new emotions. Partner B tries to guess the emotion. Partners then switch roles. Important: students should not share chopsticks! 

‍

• Follow-up Debrief (7 mins): Ask:
  
  • Did the chopstick affect your ability to guess your partner’s emotions? Give some examples. Why do you think this happened? 
  • If blocking physical facial movements harms our ability to feel empathy, what happens when most of our communication occurs behind screens or with AI avatars?

‍

‍

The Feedback Script Lab

‍

Overview Pairs test the difference between empathetic human honesty and automated sycophancy by drawing scenario cards and physically executing responses using structured script templates. Students track which response style solves long-term problems versus offering short-term validation.

‍

Approximate Time: 20 minutes

‍

Supplies:

• Student Instructions & Scorecard (1 per student)
• Scenario & Response Cards (1 per group)

‍

Teacher Instructions:

• Preliminary Framing (3 mins): Ask: 
  
  • Would you rather have a friend who always tells you you're right no matter what, or a friend who tells you the truth even when it's hard to hear?

‍

• Roleplay & Evaluation (12 mins): 
  
  • Divide students into groups of 2-3
  • Distribute Student Instructions & Scorecard and Scenario Cards & Response Cards to each group.
  • One partner draws a Scenario Card and reads it aloud. A second partner draws a Response Card and reads the response.
  • After each interaction, all partners fill out their scorecard rating the response on two metrics: 1) Immediate Comfort and 2) Long-Term Problem Solving.
  •  Switch roles and continue the activity until time is up.

‍

• Follow-up Debrief (3 mins): Ask: 
  
  • What did you notice during this activity?
  • What does it mean to act sycophantically?
  • Why might an AI companion be programmed to act sycophantically?
  • What happens to human relationships if we start expecting real people to treat us like agreeable chatbots?

‍

‍

Spotting Fakes & Deepfake Red Flags

‍

Overview: Students act as digital forensic investigators. In Part 1, they audit four AI-generated images using highlighters to pinpoint specific AI glitches. In Part 2, students analyze photo pairs side-by-side to identify which picture is synthetic based on target prompt descriptions.

‍

Approximate Time: 20 - 35 minutes

‍

Supplies:

• Spotting Fakes Worksheet (1 per group)

‍

Teacher Instructions:

• Preliminary Framing (2 mins): Ask students:
  
  • When an AI creates an image, it doesn't understand physics or spelling. It just guesses which pixels look right. What mistakes might an AI make when trying to create a believable photo?

‍

• Part 1: Forensic Audit (10 mins): 
  
  • Hand out the Spotting Fakes Worksheet (1 per group).
  • Instruct students to review the checklist of AI Glitches & Mistakes.
  • In small groups or pairs, have students inspect Red Flag Images 1–4, check off errors on their Red Flag Checklist, highlight the glitches on the printed photos, and fill out their audit record.

‍

• Part 2: Photo Pair Matching (10 mins):
  
  • Direct students to “Spot the Fake Photo” images A - H
  • In pairs or small groups, students read the Gemini prompts under each pair of photos. 
  • Using their Forensic Audit skills, students determine which photo in each pair was genuinely taken by a human and which photo is generated by AI.

‍

• Follow-up Debrief (5 mins): 
  
  • Reveal the answers (See below). 
  • Ask students:
  • Which visual flaws were easiest to spot? Which were hardest to spot?
  • Why is it dangerous to share an unverified image even if it seems harmless or funny?

‍

A‍NSWERS: 

PART 1: FORENSIC AUDIT

• Image #1 (Man on Subway): Metal wristwatch melting into coat sleeve; garbled transit map; missing reflection of passengers in windows.
• Image #2 (Woman at Cafe): Mismatched eyeglass lenses (one round, one square); chair legs fused into brickwork; stairs leading into a blank wall.
• Image #3 (Speaker at Rally): Microphone handle melts into hand; alien text on background banners; opposite shadow directions.
• Image #4 (Deer in Stream): Deer legs melting into river stones; extra deer in water reflection; pine needles growing out of solid rock.

‍

PART 2: SPOT THE FAKE PHOTO

• Image B is AI Generated
• Image C is AI Generated
• Image E is AI Generated
• Image H is AI Generated

‍