How to Play#

A simple guide to the exercise for first-time participants.

---

What Is This?#

Project Threshold is an 8-hour tabletop exercise where participants explore how AI adoption will affect major industries of the US economy over the next 2–4 years. You'll step into the shoes of an industry leader, make strategic decisions in response to evolving scenarios, and discover how those decisions ripple across the broader economy.

The goal is not to "win" — it's to build intuition about AI's economic effects and test what strategies hold up under realistic pressure.

---

Who Participates?#

There are two roles in the exercise:

Industry Participants (5–11 people). Each person represents one industry. You make decisions for your industry across four rounds. The eleven possible industries are:

Not every exercise uses all 11 industries. The facilitator selects which industries are in play based on participant count and expertise. A minimum of 5 industries ensures cross-industry richness.

Facilitator (1–2 people). Runs the exercise, reads scenario updates, scores decisions, moderates discussion, and administers the Collective Bonus (Optional) and Facilitator Market Shock mechanics.

---

Industry Selection#

Before the exercise, the facilitator assigns each participant to one industry. Ideal: 8+ industries assigned to participants. Fewer is possible — the facilitator plays unassigned industries using base case fallback decisions. If you're running with fewer than 11 participants:

Your industry assignment determines your packet, private cards, and AI adoption arc.

---

What You'll Actually Do#

Before the Exercise#

Read three documents (about 45 minutes total):

1. Executive Overview — the economic scenario and baseline forecasts
2. Your industry packet — the specific position, constraints, and profile for your industry
3. This document — so you understand the flow

You may also review your AI Adoption Arc (from AI Adoption Arcs), which shows how AI adoption unfolds round-by-round for your specific industry.

On Exercise Day#

The day follows this rhythm:

---

How a Round Works#

Every round follows the same structure and takes about 65 minutes. There are four rounds total.

---

The Decision Worksheet#

Each round, you submit one worksheet for your industry. The worksheet captures:

Don't aim for precision. The exercise uses broad bands, not exact figures. "Material but manageable investment, 3–12 month timeline, medium complexity" is exactly the right level of specificity.

---

Market Pressure Mechanics#

Two mechanics create external market pressure throughout the exercise — just like real market dynamics.

Collective Bonus (Optional, Every Round)#

During each cross-industry discussion, any participant may nominate an industry for a +2 bonus (strong strategy) or -2 penalty (risky strategy). You cannot nominate your own industry. A nomination takes effect only if 3+ participants agree. Multiple nominations may occur in a single round. If fewer than 3 participants agree, no adjustment is made.

This creates optional peer accountability and organic market pressure. Strong reasoning can be rewarded; risky reasoning can carry consequences — but only when the group reaches consensus.

Facilitator Market Shock (Round 2 Only)#

At the start of Round 2, the facilitator imposes one constraint each on 2–3 industries from the constraint menu (Regulatory Halt, Labor Cost Surge, Capital Tightening, Reputational Pressure, Competitive Response, Litigation Risk). Any industry can be hit — not just those performing poorly. This forces adaptive decision-making mid-exercise.

Industry Health Signals#

After each round (starting Round 2), the facilitator announces how your industry is doing based on your cumulative score across all rounds. There are five conditions:

These conditions create real consequences — strong early decisions build momentum, while poor decisions compound into structural disadvantage.

---

Private Information Cards#

In Rounds 1–3, you receive a confidential card with industry-specific intelligence (from Private Cards). This might be internal data (e.g., your AI pilot results were weaker than expected), competitive intelligence (e.g., a competitor is about to launch a disruptive product), or regulatory signals (e.g., a major enforcement action is coming).

You can reference your card's information in your decisions and in discussion, but you shouldn't read the card aloud verbatim. Cards accumulate — keep them all and use them in later rounds.

There is no card in Round 4. The final round operates on public information only.

---

AI Adoption Arc#

Each industry has an AI Adoption Arc (from AI Adoption Arcs) that describes how AI adoption unfolds round-by-round for your specific industry. This provides context for your decisions: what stage of AI deployment your industry is in, what capabilities are becoming available, and what constraints are shifting.

Review your arc before the exercise and reference it each round as you make decisions.

---

Scoring#

Decisions are scored on three dimensions, each rated -2, 0, or +2:

A score of +6 means you nailed all three. A score of 0 means your decision was safe but unremarkable. Negative scores mean your logic had gaps or you ignored real constraints.

Scoring exists to create structure and feedback — it's a tool for learning, not a competition.

---

The Debrief (60 min)#

The debrief is the most important part of the exercise. It is non-negotiable and cannot be shortened.

---

Key Principles#

You're building intuition, not being tested. There are no right answers. The value is in the reasoning, the trade-offs you surface, and the cross-industry dynamics you discover.

Decisions are binding. Once you submit a decision, you can't undo it. You can modify or build on it in later rounds, but you live with your choices. This mirrors real strategic decision-making.

The discussion is where the learning happens. The 27-minute cross-industry discussion each round is not filler — it's the most valuable part of the exercise. Challenge each other. Ask "how does your decision affect my industry?" Surface the interdependencies.

Use broad strokes, not false precision. When describing your decisions, use the provided bands (low/medium/high, pilot/regional/national). Don't invent specific dollar amounts or percentages unless the exercise materials provide them.

Private information creates realistic asymmetry. Not everyone knows the same things. This is by design. It mirrors real markets where different actors have different intelligence.

---

Quick Reference: The Eleven Industries#

8-Hour Session Timeline#

Total Duration: 8 hours exactly (e.g., 8:30 AM to 4:30 PM) Key Principle: Debrief is protected and non-negotiable. All flexibility built into rounds and breaks.

---

EXERCISE TIMELINE#

---

ROUND STRUCTURE (Rounds 1–4) — INDIVIDUAL INDUSTRY PARTICIPANT MODEL (V7.4)#

Each round follows this sequence. Total per round: 1:05. Each industry participant submits one decision per round, with cross-industry discussion emphasis.

Facilitator Notes for Rounds (V7.4 Industry Model):

• Round 1 (9:10–10:15): Baseline round. Participants may be cautious. All industries expected to submit decisions (individuals decide quickly). First Collective Bonus opportunity establishes early group consensus on strategy quality. Facilitator highlights key spillovers between decisions.
• Round 2 (10:30–11:35): Acceleration. Facilitator Market Shock: Before decision prep, facilitator selects 2–3 industries and imposes one constraint each from the constraint menu (~3 min). Inject creates urgency. Cross-industry discussion reveals how constraints cascade across industries (e.g., Consulting pricing pressure affects Finance advisory spend; Law AI adoption creates precedent concerns). Collective Bonus continues.
• Round 3 (12:15–1:20): Reckoning. External shock injects. Collective Bonus carries heightened stakes as cumulative scores diverge. Participants adjust strategy based on peer signals and Industry Health conditions.
• Round 4 (1:35–2:40): Normalization. No private cards; full transparency. Final decisions for long-term positioning. Final Collective Bonus opportunity. Cross-industry discussion focuses on emergent patterns and no-regrets actions. Expect all industries to submit.

---

DEBRIEF STRUCTURE (60 minutes, 2:50–3:50)#

This block is non-negotiable and protected from round overruns.

---

BUFFER & FLEXIBILITY NOTES#

Scaling Note for 11 Participants: With more industry participants, scoring time may need slight extension. If running 11 industries, facilitator may need a second scorer or simplified scoring approach during the 12-minute scoring block. Plan for approximately 1 minute per industry decision.

---

FACILITATOR CHECKLIST#

• 8:25 AM: All materials printed and organized by industry
• 8:25 AM: Scoring rubric and Collective Bonus procedure reviewed with all facilitators
• 9:00 AM: Confirm all industry participants have decision submission forms + pens
• 9:10 AM: Begin micro-round; watch for timing
• Each round: Set visible timer (countdown clock projected if possible)
• Each round: Call "3 minutes" warning for individual prep hard stop
• Each round: Distribute private cards from Private Cards at correct time (Rounds 1–3 only)
• Post-round: Facilitators score within 12 min; announce Industry Health conditions
• 2:50 PM: Transition to debrief; do not allow round content to bleed into debrief time
• 3:50 PM: Hard stop debrief; distribute feedback surveys
• 4:25 PM: Thank industry participants; remind about synthesis memo due date (within 5 business days)

---

TIME-AT-A-GLANCE#

08:30 +- Arrival
08:45 +- Welcome
09:00 +- Micro-Round
09:10 +- ROUND 1 (Foundation)
10:15 +- Break
10:30 +- ROUND 2 (Acceleration)
11:35 +- Lunch
12:15 +- ROUND 3 (Reckoning)
01:20 +- Break
01:35 +- ROUND 4 (Normalization)
02:40 +- Break
02:50 +- DEBRIEF (60 min, non-negotiable)
03:50 +- Wrap-Up
04:30 +-- END

Total: 8 hours exactly. Debrief protected. Ready to execute.

Practice Round Walkthrough#

Purpose#

This document scripts a complete practice micro-round so that all individual industry representatives understand the decision-making flow, scoring mechanics, and submission process before Round 1 begins. The facilitator can read this walkthrough aloud, adapt it to fit timing, and answer questions that arise during execution.

Key Goal: Individual participants will have confidence in the worksheet format, understand how scoring works, and know what to expect when time pressure increases in the live rounds.

---

Overview#

• Duration: 10 minutes total (scripted, compressed from live rounds)
• Materials: 5-11 individual industry representatives (each with 1 blank decision worksheet)
• Scenario: Simple, AI adoption in a professional services environment
• Output: Minimum 5 individual decisions (1 per industry participant)
• Key Change: Each individual industry representative submits ONE decision per round. If a participant is absent or passes, they receive a pre-defined base case fallback (see Base Case Fallback Bank).

---

Facilitator Script: Setup (1 minute)#

"Good morning, everyone. Before we launch Round 1, we're going to run a quick practice round together. This is a sandbox — nothing here counts toward your final score. The purpose is threefold: first, you'll see the exact worksheet format; second, you'll watch scoring happen in real time; and third, you'll practice the rhythm of individual decision preparation, submission, and feedback.

We're going to compress the timing today — 10 minutes instead of 15 minutes of decision prep. That means you'll move fast. That's intentional. In the real rounds, you'll have more time to think through risk, but the structure is identical.

Each of you will decide individually for your industry. No team deliberation — you analyze and decide on your own.

Everyone clear? Good. Let's begin."

---

Step 1: Situation Update (1 minute)#

The Scenario#

"It's January 2026. A major AI company has just released a new foundation model that scores 90% on professional licensing exams — think bar exam, medical boards, engineering exams, you name it. Adoption is ramping fast across industries. Your industry needs to decide how to respond in the next six months.

Here's your inject card — this is new information that just arrived."

[Facilitator reads aloud]

INJECT: Three Fortune 500 companies announce they are cutting entry-level hiring by 25% and redirecting savings to AI tool licenses.

"This means the labor market is tightening. Fewer junior roles. More pressure to upskill existing staff with AI. Every industry sees it — but it hits some harder than others.

Each of you will now submit one decision for your industry. That decision describes what you'll invest in, how long it takes, what it costs, and what success looks like.

Let's go."

---

Step 2: Individual Decision Preparation & Submission (3 minutes)#

Individual Instructions (read aloud)#

"Each of you has three minutes to analyze the situation and complete ONE decision worksheet for your industry. You are deciding alone — no team deliberation.

Key rule: When you submit a decision, your industry gets scored. If you decide to skip or don't submit, your industry receives a pre-defined base case fallback (see the Base Case Fallback Bank reference). That fallback scores 0 on all dimensions but keeps your industry in play.

Think critically about your industry's position. What's your response to the hiring freeze inject? What AI investments make sense given the constraint?

You must decide: What's your industry doing in the next 6 months?

Time starts now. Go."

---

Example Decisions (Script If Needed)#

If a participant is stuck, the facilitator can offer these examples to unblock them. NOTE: These examples use BANDED FRAMEWORK instead of granular costs.

---

Worksheet Template (Show One Completed Example)#

[Facilitator shows the Retail decision on screen or poster.]

"Here's what a filled-out worksheet looks like. Notice the key fields — and remember, we're using BANDS instead of invented precision:

• Industry at the top
• Decision Title — plain language, clear scope
• Description — the 'who, what, where, when' (directional, not granular)
• Timeline — milestones or phases (avoid false precision like 'month 1.5')
• Spend/Commitment Band — {Absorbable (BAU), Material but manageable, Transformational bet, Potentially existential}
• Time-to-Impact Band — {0-3 months, 3-12 months, 1-2 years, 2+ years}
• Execution Complexity Band — {Low, Medium, High, Very high}
• Dependency Band — {Mostly internal, Requires vendor/partner, Requires regulator/union/standards, Requires ecosystem shift}
• Scale Band — {Pilot / Regional / National / Global}
• Success Metric Band — {Low/Moderate/High/Transformational impact}
• Key Risk & Mitigation — be specific about the top 2-3 risks

Rules: Avoid line-item cost breakdowns. Avoid invented precision (don't say '$15.3M'; say 'Material but manageable'). Use bands. Let the band language do the work.

Important: You must submit ONE decision for your industry. If you don't submit, your industry auto-defaults to base case fallback (scores 0 on all dimensions).

You have three minutes. Fill the worksheet. Go."

[Time: 3 minutes elapsed. 4 minutes total.]

---

Step 3: Share-Out and Plausibility Challenge (2 minutes)#

Two Participants Present#

"Let's hear from two participants. Retail, you go first — 30 seconds."

Retail Participant: "Retail is piloting AI inventory optimization in 50 Midwest stores. Six-month timeline, material but manageable investment, goal is reduce markdowns and improve inventory turns. Phased rollout with vendor partnership."

Facilitator: "Good. Clear scope. Regional pilot — that's smart. Bounded risk. Next: Consulting."

Consulting Participant: "Consulting is deploying an AI copilot to all consultants — three-month rollout, material but manageable investment, goal is 15% increase in billable hours per consultant. Phased across three batches with buddy system and weekly tracking."

Facilitator Challenge#

"Consulting, I want to probe the copilot rollout. Three months to full rollout across all consultants — that's aggressive. How do you handle change management? Adoption risk?"

Consulting Participant: "Good catch. We'd phase it — 500 consultants in month 1, another 500 in month 2, full 2000 by month 3. Each batch gets two hours of e-learning, plus a buddy system pairing early adopters with skeptics. We track utilization weekly and adjust training if adoption lags."

Facilitator: "Acceptable. That's a real rollout plan, not a hope. In live rounds, I might ask: What's your fallback if only 30% of consultants adopt by month 2? But for now, this is solid. I'd score this as manageable execution risk."

[Time: 2 minutes elapsed. 6 minutes total.]

---

Step 4: Scoring Example (2 minutes)#

Scoring Mechanics#

"Now let's score two of these decisions so you see how banding works. Every decision gets scored on three dimensions — Strategic Fit, Execution Risk, and Tail Risk. Each dimension can be +2, 0, or -2. Red-flag trigger conditions can shift scores by +/-3. Let me show you."

---

Example 1: Retail — Inventory Optimization (Strong Decision)#

[Facilitator writes on board or displays:]

Facilitator commentary: "Notice what makes this strong: bounded scope (Regional, not Global), realistic complexity (Medium, not Very High), managed downside. Retail isn't betting the company. Test, learn, scale. That's strategic discipline. The banded language lets us focus on the real trade-offs instead of arguing about whether cost is $15M or $17M."

---

Example 2: Consulting — Copilot Rollout (Acceptable Decision)#

[Facilitator writes on board or displays:]

Facilitator commentary: "This is the difference between +2/+2/+2 and +2/0/0. Consulting moved boldly, which is great. But they've put execution risk management in the medium band. In a real round, I'd push: 'If only 50% of consultants adopt by month 2 instead of your assumed 70%, what happens to revenue?' That's the question they haven't fully answered. For this practice round, acceptable. For live rounds, they need to either de-risk the execution or acknowledge the tail risk more explicitly."

---

Example 3: Red Flag Trigger (Illustrative)#

"Now, let me show you what triggers a red flag. Imagine the Consulting participant had proposed: 'Fire 40% of associate staff and replace them with AI copilots. Timeline: 6 months. Cost: $2M in severance. Expected savings: $50M. Net benefit: $48M.'

That would trigger multiple flags:

• Plausibility: Is a 40% headcount reduction actually executable in six months? Legally? Operationally?
• Tail Risk: If the copilots underperform or adoption stalls, you've permanently lost institutional knowledge and client relationships. That's a tail risk that could impair revenue for years.
• Ethical/Stakeholder: Are there labor law, client service, or reputational risks you haven't mentioned?

If I see a proposal like that, I would challenge it hard. I might score it -3 on Execution Risk and -3 on Tail Risk because the downside far outweighs the upside. Or I might reject it as implausible and ask for a revised proposal.

That's how red flags work. They're rare. But they're circuit-breakers for decisions that look too good to be true or too risky to stomach."

[Time: 2 minutes elapsed. 8 minutes total.]

---

Step 5: Wrap-Up and Key Reminders (1 minute)#

Facilitator: "That's the process. Every round in Project Threshold follows this exact flow:

1. Situation update and inject.
2. Each participant submits ONE decision for their industry (minimum 5 total; up to 11 if all industries are represented).
3. Industries without explicit decisions receive pre-defined base case fallbacks from the fallback bank.
4. Two participants present; I challenge one decision for plausibility.
5. I score all submitted decisions on the three dimensions using banded framework.
6. Cross-industry discussion and optional Collective Bonus — any participant may nominate an industry for +2 (strong) or -2 (risky) if 3+ participants agree.
7. We review scores and any Collective Bonus results, then talk through the reasoning.

In live rounds, deliberation time will be 15 minutes instead of 3 — so you'll have room to stress-test your decisions, model risks, and think through cross-industry implications. That said, the worksheet format and scoring logic are identical to what you just did.

A few key reminders for Rounds 1, 2, and 3:

• Hard deadline: When I call time, submission closes. No late arrivals.
• One decision per participant: Each participant submits one decision for their industry.
• Base case fallbacks: Industries without explicit decisions auto-default to base case fallback (scores 0 on all dimensions).
• Banded framework: Use bands — avoid invented precision. 'Material but manageable' not '$15.3M'. Let the band language do the work.
• Collective Bonus is open and optional: Nominations are stated publicly. A nomination takes effect only if 3+ participants agree. This creates real but consensus-driven accountability.
• Scoring is consistent: I use the same banding logic every round. {-2, 0, +2} per dimension, with red-flag triggers at +/-3.
• Red flags are rare and explicit: If I flag a decision, I'll tell you exactly what I'm concerned about. You can then decide to revise or defend.

Any questions?"

[Allow 30 seconds for Q&A.]

Facilitator: "Excellent. You're ready for Round 1. We'll begin in five minutes. Grab water, stretch, and reconvene. See you shortly."

[Time: 1 minute elapsed. 9 minutes total. Total time: ~10 minutes.]

---

Appendix: Blank Decision Worksheet#

The Decision Worksheet is a standalone printable form. See Decision Worksheet for the full template.

Print multiple copies per participant before the exercise begins. Each participant submits one worksheet per decision per round.

Quick reference — the 12 required fields:

---

End of Practice Round Walkthrough#