🔒 Exercise 2 Facilitator Guide

Password-Protected Instructor Material

Exercise 2: AI-Assisted Strategic Analysis - Facilitator Guide

Duration: 70 minutes (11:20 AM - 12:30 PM) Core Learning: AI as strategic thought partner. Best results: Human expertise + AI analysis.

Pre-Exercise Setup (During 10:30 Break)

Technology: - Have ChatGPT/Claude open on YOUR laptop (backup if tables lack devices) - Recommend: Google Gemini (free, no login) - Test venue wifi—this is critical

Room: - Flipcharts at each table with Three Horizons grid pre-drawn - RED sticky notes (threats), GREEN (opportunities), BLUE/YELLOW (AI insights) - Strategic Prompting Guide handout at each seat

Announcement (11:00): “Next exercise uses AI tools. Each table: designate one ‘AI Lead’ who types prompts. You’ll try three techniques: 1. Risk Deep-Dive: AI identifies threats you missed 2. Reverse Prompting: AI asks YOU questions 3. AI Debate: Two AIs argue different perspectives

Google Gemini recommended (free, no login). If your table lacks devices, I’ll demonstrate live.”

Phase 1: Human Brainstorm (15 min)

Your intro (2 min): “Phase 1 is YOU, no AI yet. 15 minutes.

RED sticky notes: AI threats to your business GREEN sticky notes: AI opportunities

Place on Three Horizons grid. One idea per note. Go.”

During (10 min): - Circulate, ensure all tables generating ideas - If stuck: “What keeps you up at night about AI?” or “Where do you waste human time AI could help?” - Don’t give ideas, don’t critique yet

Common patterns (expected, not wrong): - Most threats H1-H2 (obvious) - Most opportunities H3 (aspirational) - Clustering (many similar ideas)

Phase 2: AI Augmentation (30 min)

Setup (1 min): “Three techniques. Tables: Open ChatGPT/Claude/Gemini. Copy prompts from your guide.”

Round 1: Risk Deep-Dive (10 min)

Your script: “Use the Risk Deep-Dive prompt. Customize with YOUR threats. Submit. Read aloud to your table.”

During (8 min): - Circulate rapidly - Listen for “aha!” moments - If silent: “What did AI identify that you missed?” - If dismissive (“generic”): “Try a follow-up: ‘Be specific to [your industry]’” - If resistant (“AI doesn’t know our business”): “Right. But what kernel of insight is useful?”

Common “aha!”: - “AI connected dots we didn’t (pricing → data flywheel)” - “AI brought adjacent industry examples” - “Second-order effects we missed”

Round 2: Reverse Prompting (10 min)

Your script: “Instead of asking AI for answers, ask AI to ask YOU questions. This is powerful.

Submit the Reverse Prompting prompt. Discuss AI’s questions as a table 1-2 min BEFORE typing your answer. Then respond. AI asks next question.”

During (8 min): - Watch for “aha!” moments (“That question reveals we’re overlooking X!”) - If going too fast: “Don’t rush. Discuss the questions.” - If stuck: “What did that question make you realise?”

Common discovery: - Questions more valuable than answers - Reveals hidden assumptions - “We’ve never questioned why we do this”

Round 3: AI Debate (10 min)

Your script: “Final technique: AI Debate. Show of hands: Should we invest in AI NOW despite data limitations? Or wait for data infrastructure first?

I’ll demonstrate. Watch the AIs argue this out.”

On projector: 1. Submit “AI Optimist” prompt (ChatGPT) → Read opening aloud 2. Submit “AI Pragmatist” prompt (Claude) → Read opening aloud 3. Copy Optimist’s argument → paste into Pragmatist → Read counter 4. Copy Pragmatist counter → paste into Optimist → Read counter 5. Let debate 2-3 rounds

After debate: “Who finds Optimist more convincing?” [Hands] “Pragmatist?” [Hands] “Why? What made that argument stronger?”

Key point to surface: “Usually the answer is ‘do BOTH’—pilot now, build infrastructure in parallel. Debates help you see synthesis.”

Phase 3: Human Synthesis (15 min)

Framing (2 min): “What did you learn? Not ‘what did AI say’ but ‘how did AI change YOUR thinking?’”

Table discussion (5 min): “Discuss: 1. What did AI identify that humans missed? 2. What did humans see that AI missed? 3. Which AI insights are most valuable? 4. How would you use this going forward?”

Your role during discussion: - If focusing only on AI’s value: “What did AI MISS? What context did you have?” - If dismissing AI: “Did AI identify ANY insight worth considering?” - If debating “was AI right”: “AI isn’t right or wrong—it’s a thought partner. Did it make you think differently?”

Capture themes on flipchart: - “AI has breadth, we have depth” - “AI challenges assumptions we didn’t know we had” - “AI helps, humans still decide”

Shareout (8 min): “Each table: Share ONE AI insight that changed your thinking.”

Common Challenges & Fixes

Challenge	Fix
“AI gave generic responses”	“Try adding context: industry, specific challenge, ‘be concrete’”
“Table has no internet”	“Watch my screen—we’ll demo together”
“This feels forced”	After Phase 2: “Did AI add ANY value?” Usually yes. “That’s why.”
“Going too fast/slow”	Fast: Add 4th technique or deepen discussion. Slow: Skip AI Debate, you summarize.
“Students want AI for everything”	“AI helps you THINK. Frameworks help you DECIDE. You need both.”

BYOD Troubleshooting

Multiple laptops at table: “Your choice—everyone on own device or huddle around one. Whatever helps you think.”

No devices at table: “Watch my screen. Demo works here. Learning still happens.”

Slow wifi: “While you wait, discuss: What do you EXPECT AI to say? Then see if you’re right.”

AI says something wrong: “Perfect catch. This is why you evaluate critically. AI generates, you validate.”

Success Indicators

✅ Exercise worked if: - Tables have “aha!” moments during exercise - Someone uses AI during Dragon’s Den prep without you suggesting it - Debrief includes: “AI helped me see X I missed” - Students cite “reverse prompting revealed…” in final reflection

Post-Exercise

Immediately after (12:30): - Note which tables had great AI interactions - Note which struggled (tech issues? poor prompts? skepticism?)

Afternoon Dragon’s Den setup (1:45): Say casually: “AI tools available if teams want to use them during prep. You practiced Risk Deep-Dive, Reverse Prompting, AI Debate this morning. Use any technique that helps. Or don’t.”

Then let it happen organically. Don’t mandate.

Your Checklist

Before: - [ ] Test ChatGPT/Claude on your device - [ ] Wifi works - [ ] Flipcharts ready with Three Horizons grid - [ ] Sticky notes distributed - [ ] Strategic Prompting Guide at each seat

During: - [ ] All tables submit prompts (or watch your demo) - [ ] Listen for “aha!” moments, amplify them - [ ] Model good prompting if tables struggle - [ ] Manage time—strict phase transitions

After: - [ ] Note what worked, what didn’t - [ ] Available for questions during lunch - [ ] Casual mention of AI availability this afternoon

Flexible, responsive facilitation. Key is: Let them experience AI as thought partner, not magic.