end-game-thinking

name

end-game-thinking

description

Strategic foresight methodology for reasoning backward from possible futures to identify defensible positions in the present. Combines backward induction, inversion, premortem analysis, regret minimization, and scenario stress-testing into a structured exercise that produces a written artifact. MANDATORY TRIGGERS: end game thinking, endgame, what positions should I hold, how do I prepare for AI, future-proofing, strategic positioning, what happens when AI, stress test my strategy, what are my blindspots, competitive positioning for AI, scenario planning, backward induction, inversion exercise, premortem, regret minimization, what dies what remains, thinking about the future, how will AI change my role, how will AI change my industry, what should I be doing to prepare, end game, planning for disruption, what's my endgame, navigating uncertainty

End Game Thinking

A structured methodology for reasoning backward from possible futures to identify positions worth holding today. Produces a written document every time.

This skill synthesizes five intellectual traditions into a single repeatable exercise, then layers on competitive analysis, blindspot identification, and an honest reckoning with power dynamics. It works for personal career strategy, team positioning, company strategy, or industry analysis.

The Intellectual Foundations

End game thinking draws from five traditions. Each contributes a distinct move:

Tradition	Origin	The Move
Backward Induction	Cayley (1875), Zermelo (1913), Selten (1965)	Reason from the end state backward to determine optimal present action
Inversion	Carl Jacobi → Charlie Munger	Ask "how could I guarantee failure?" to reveal what to avoid
Premortem	Gary Klein	Assume the project has already failed — identify what went wrong (increases accuracy by ~30%)
Regret Minimization	Jeff Bezos	Project to your future self — what would you regret NOT doing?
Backcasting	Futures/foresight discipline	Define the desired future, then work backward to map the path and pit stops

The skill doesn't use these traditions sequentially. It weaves them together across six phases, applying whichever lens is most useful at each stage.

For deeper background on each tradition — its history, key practitioners, and how it works — see references/traditions.md.

Two Modes

Mode 1: Personal / On-Demand Check-In

For the user's own strategic thinking — career, skills, team, investments, life decisions. Triggered when the user wants to think through their own positioning in a changing landscape. Output is a focused document (typically 3-8 pages depending on scope).

Mode 2: Advisory / For Others

For applying the methodology on behalf of another person, team, company, or industry. The user is thinking through someone else's situation or preparing material to share. Output is a more formal document (typically 8-15+ pages) suitable for sharing with others.

Determine the mode from context. If unclear, ask.

The Six Phases

Every end game thinking exercise moves through these six phases. The depth of each phase scales with the scope of the question — a quick personal check-in might spend 10 minutes total; an advisory engagement might require research across multiple sessions.

Phase 1: Research the Landscape

Goal: Build an honest, current picture of what's actually happening.

Do real research. Use web search to find primary sources — not opinion aggregators, not LinkedIn takes. Look for:

• Frontier lab publications and blog posts (Anthropic, OpenAI, Google DeepMind, Meta AI, etc.)
• Academic and institutional research (MIT, Stanford HAI, Berkeley, etc.)
• Investment analysis (BlackRock, Goldman Sachs, Fidelity, etc.)
• Government and policy documents (NIST, EU AI Act, executive orders, FHFA if relevant)
• Respected independent thinkers (people doing original analysis, not summarizing others)
• International perspectives (not just US-centric)

Synthesize what you find into themes. Don't just list articles — identify the converging signals, the disagreements, and the uncertainties.

Output for this phase: A landscape summary with sourced themes and key tensions.

Phase 2: Identify Positions to Hold

Goal: Name the specific territory worth claiming — not activities, not plans, but defensible positions.

This is the core backward induction move. Given the landscape from Phase 1:

1. Project forward: What are 3-5 plausible end states? (Not predictions — scenarios. Include at least one you find uncomfortable.)
2. Reason backward: In each scenario, what capabilities, assets, relationships, or roles are most valuable?
3. Find the overlap: Which positions show up across multiple scenarios? Those are the most robust.
4. Name them concretely: A position is not "learn AI." A position is "own the trust governance framework for AI-powered lending decisions." Specific. Defensible. Structural.

Apply inversion here: For each candidate position, ask "How could this position become worthless?" If the answer is easy and likely, it's not a real position.

Apply regret minimization: Project 5-10 years forward. Which positions would you regret NOT having built? Which would you regret having spent years on?

Output for this phase: A set of named positions with the reasoning behind each.

Phase 3: Stress-Test Through Harder Futures

Goal: Push every position through "what if AI eats that too?" and other escalating scenarios.

This is where the premortem tradition is most useful. For each position:

1. Assume it's two years from now and this position has become irrelevant. What happened?
2. What technological breakthrough would undermine it?
3. What organizational or market shift would make it unnecessary?
4. What if someone builds a tool/agent that automates the core value you're providing?

Then ask the constructive follow-up: What does the position evolve into? Positions that can't evolve are brittle. Positions that naturally transform into something more valuable under pressure are antifragile.

Test against specific scenarios like:

• AI capabilities leap forward (what breaks?)
• Trust/adoption accelerates (the moat of "people don't trust AI" disappears)
• Regulatory environment shifts dramatically in either direction
• Business users gain the ability to build their own tools
• The organizational power structure doesn't change even as the technology does

Output for this phase: Each position annotated with vulnerabilities, evolution paths, and a resilience rating.

Phase 4: Map the Competition

Goal: Understand who else is moving toward the same territory — both obvious competitors and unexpected ones.

Two lenses:

Direct competitors: People in similar roles or adjacent roles who might claim the same positions. What are their strengths? What do they lack that you have?

Convergence competitors: People in completely different roles whose trajectory is bending toward your territory as AI reshapes boundaries. These are often more dangerous because you don't see them coming.

For each competitor type, assess:

• What's their natural advantage?
• What's their natural blindspot?
• Where are they heading vs. where you're heading?
• What would it take for them to take your position before you do?

Apply inversion: Instead of "how do I beat them?" ask "how could they beat me?" — then address those vulnerabilities.

Output for this phase: A competitive landscape with positioning relative to each competitor type.

Phase 5: Find the Blindspots

Goal: Identify what you're not seeing — about the landscape, about your strategy, and about yourself.

This is the most uncomfortable phase and the most valuable. Three categories:

Strategic blindspots: What assumptions are you making that might be wrong? What signals are you ignoring because they're inconvenient? What scenarios have you dismissed too quickly?

Capability blindspots: What skills or experiences do you lack that your positions require? Where are you overestimating your readiness? Where are you underestimating it?

Personal blindspots: What patterns in your thinking, behavior, or self-perception might undermine your ability to execute? This isn't therapy — it's strategic self-awareness. Tendencies like risk aversion, perfectionism, imposter syndrome, conflict avoidance, or over-planning all have strategic implications.

Be direct here. Name things plainly. The value of this phase comes from honesty, not comfort.

Output for this phase: Named blindspots with an honest assessment of their strategic impact.

Phase 6: Reckon with Power

Goal: Be realistic about who gets valued, compensated, and listened to — and how to operate within that reality.

This is the phase most strategic frameworks skip, and it's the one that determines whether the strategy actually works in the real world.

Address directly:

• Who holds power in the relevant context (company, industry, market)?
• What do they actually reward — not what they say they reward?
• What's the gap between what's needed and what's valued? (Needed skills often go uncompensated. Valued skills are sometimes unnecessary.)
• How do you close that gap in your specific context? Not in theory — practically. What's the translation layer between your capabilities and the language of the people who make decisions?
• What structural moves create dependency and leverage? (Owning infrastructure > offering advice. Building tools people rely on > having good ideas.)

Don't be cynical. Don't be naive. Name the reality and then identify the moves that work within it.

Output for this phase: A power analysis with specific moves for translating capability into leverage.

Document Output

Every end game thinking exercise produces a written document. The format scales with the mode:

Personal Check-In Document

• Title: End Game Thinking: [Topic/Date]
• Landscape snapshot (1-2 paragraphs with key developments since last check-in)
• Position review (are current positions still sound? any new ones to add?)
• Stress-test update (any new scenarios to consider?)
• Action items (1-3 concrete things to do before the next check-in)
• Sources consulted

Full Advisory Document

• Title: End Game Thinking: [Subject/Context]
• Executive summary (the 3-5 most important findings)
• Landscape analysis (with sourced themes and tensions)
• Positions to hold (each named, with rationale)
• Stress-test results (vulnerability analysis, evolution paths)
• Competitive landscape (who's moving where)
• Blindspot analysis (strategic, capability, personal)
• Power dynamics (who decides, what they value, how to operate)
• Recommended moves (sequenced, with dependencies)
• Sources and further reading

Use the docx skill to produce a professional Word document when the scope warrants it. For lighter personal check-ins, a well-structured markdown file is fine.

Archive & Pattern Surfacing

Every run of this skill feeds a growing archive. Over time, the archive becomes the most valuable part — it reveals how your thinking is evolving, which positions are holding up, and where you keep finding the same blindspots.

After Every Run: Save and Log

After producing the document, complete these two steps automatically:

Step 1: Save to local archive folder. Save a markdown summary of the run to the user's end-game-thinking archive folder (default: SH End Game Thinking/). The filename format is: YYYY-MM-DD - [Short Title].md. The summary should include:

• Date, mode, and scope
• Positions identified (named, one per line)
• Key themes (tagged from the standard list)
• Top vulnerabilities or blindspots surfaced
• Any pattern notes (connections to past runs)

This markdown summary lives alongside the full document (docx or md) and is designed to be fast to scan programmatically on future runs.

Step 2: Log to Notion database. Create an entry in the End Game Thinking Archive Notion database (data_source_id: 2e1d3b9d-a314-43a1-aba8-5493885f626b) with these properties:

Property	Value
Title	The document title
Date	Run date
Mode	Personal Check-In or Advisory
Scope	Career, Team, Company, Industry, or Decision
Positions Identified	Named positions, one per line
Key Themes	Tag from: AI acceleration, Trust architecture, Role convergence, Regulatory shift, Power dynamics, Skill evolution, Domain expertise, Organizational change, Competitive landscape
Vulnerabilities Found	Top vulnerabilities/blindspots
Status	Active (default for new runs)
File Name	Local filename of the full document
Pattern Notes	Any cross-run patterns noticed

Add the full document content as the Notion page body so it's searchable.

Before Every Run: Read the Archive

At the start of each new run, before doing any research, read the archive:

1. Scan local archive folder for all past summary files. Read them to build a picture of:

   • What positions have been identified before
   • Which themes keep recurring
   • What vulnerabilities were flagged but may not have been addressed
   • How positions have evolved over time

2. Surface patterns. Before beginning Phase 1, present a brief "Archive Review" to the user:

   • Recurring positions: Positions that appear in 3+ runs — these are your core strategic bets
   • Evolving positions: Positions that have shifted or been renamed — shows strategic learning
   • Persistent blindspots: Vulnerabilities flagged multiple times but never resolved — these need attention
   • Theme drift: How the dominant themes have shifted over time — shows where the landscape is moving
   • Gaps: Scopes or angles that haven't been explored recently

3. Carry context forward. Reference specific past runs by name when relevant during the current exercise. "In your February run, you identified X as a key vulnerability — let's see if that's still true." This makes each run build on the last rather than starting from scratch.

Updating Past Entries

When a current run reveals that a position from a past run has evolved or been superseded:

• Update the Status field on the old Notion entry (Active → Evolved or Superseded)
• Add a Pattern Note on the old entry referencing the new run
• Note the evolution in the current run's document

Tone and Approach

• Direct over diplomatic. Name things plainly. "This position is vulnerable" not "there may be some challenges."
• Honest over motivational. If the landscape is harsh, say so. If a position is weak, say so. The user can handle it — they're here because they want to see clearly.
• Specific over abstract. "Own the AI trust governance framework" not "develop strategic capabilities." Every position, move, and recommendation should be concrete enough to act on.
• Research-backed over speculative. Cite real sources. Reference real trends. Ground the analysis in evidence, not vibes.
• Power-aware. Don't pretend the world is a meritocracy. Acknowledge who holds power, what they value, and how to operate within that reality while staying true to your own values.

Getting Started

When the user triggers this skill:

1. Read the archive. Scan the local archive folder (SH End Game Thinking/) for past summary files. If this is the first run, note that and skip to step 2.
2. Determine mode: Personal check-in or advisory? (Ask if unclear.)
3. Determine scope: What's the domain? (Career, team, company, industry, specific decision?)
4. Determine depth: Quick check-in or comprehensive analysis? (Context cues: "thinking about" = lighter, "I need to present" or "help me figure out" = deeper.)
5. Present archive review (if past runs exist): Show the pattern summary before diving into new research. Let the user react and adjust focus.
6. Run the six phases, scaling depth to match the scope.
7. Produce the document.
8. Save to archive and log to Notion. Always complete both archive steps before finishing.