Akili King

Multi-Domain Problem Solving

30 Years of Boyd-Based Problem Solving

Solving Complex Problems
Across Domains

Boyd's strategic theories (1994) + 23 years military + 15 years systematic study + LLMs as latest tool = 30-year proven methodology for complex operational challenges.

A 30-Year Methodology

This approach to complex problem-solving was built over three decades, not invented with LLMs.

1994
Boyd Foundation
Introduced to John Boyd's strategic theories at West Point
1991-2014
Military Service
23 years applying Boyd's principles in combat and intelligence
2010-2022
Systematic Study
12+ years curated reading: Drucker, Lean, complexity
2018-2022
AI Study
4+ years AI/ML study BEFORE ChatGPT existed
2022-Now
LLM Tools
Latest tool in 30-year methodology

The Core Methodology: Boyd Since 1994

John Boyd's "Destruction and Creation," "Patterns of Conflict," and "Strategic Game" have shaped my approach to complex problems for 30 years—since West Point in 1994. LLMs are just the newest tool for applying these time-tested principles.

ChatGPT launched November 2022. My problem-solving foundation was established 28 years earlier.

The Boyd Method

How I Work: The Strategic Game

Since 1994, I've applied John Boyd's "Destruction and Creation" methodology to solve complex problems across domains. This is not theory—it's a 30-year practice of breaking apart concepts and synthesizing new solutions from the pieces.

The Four-Step Strategic Game

Boyd's approach asks: What lies hidden under the question marks? To find the answer, you must examine problems from multiple perspectives and synthesize patterns across domains.

1
General Survey

Examine the problem from multiple perspectives. Look across different domains—not just within one field. Ask: What is strategy? What is the aim? What are the key ideas?

2
Condense to Essential Elements

Break apart concepts into constituent parts (Destructive Deduction). Identify the core components that matter. Strip away the non-essential.

3
Place in Strategic Perspective

Synthesize patterns ACROSS domains (Creative Induction). Find common qualities and link pieces together in new ways. Test for consistency and match with reality.

4
Implementation

Apply the synthesized solution in real environment. Measure outcomes. Feed results back into the cycle. Repeat: Survey → Condense → Synthesize → Implement.

Three Complementary Frameworks

Boyd's 4-step Strategic Game is strengthened by three complementary frameworks that address different aspects of organizational complexity:

Orlikowski's Duality of Technology (1992)

Technology is BOTH the product of human action AND the medium that structures human action. This duality explains why technology-first approaches fail: deploying AI tools alone doesn't work because technology must be appropriated by users within their organizational context.

Applied in 4-step process: Step 1 examines organizational structures and human agency. Step 2 identifies how practices enable/constrain adoption. Step 3 designs technology appropriation fitting organizational context. Step 4 recognizes users reconstruct technology through use.
Cynefin / Anthro-Complexity (Snowden)

The "science of common sense" for beings who "are not ants, or birds." Humans are homo-faber (makers), homo-ludens (players), and homo-narrans (storytellers). The 3Is—Intelligence (reflection, abstraction), Intentionality (purpose, choice), Identity (how we "show up")—cannot be reduced to computational models.

Applied in 4-step process: Step 1 identifies problem domain (complex vs complicated). Step 2 focuses on the 3Is. Step 3 applies appropriate sense-making. Step 4 uses probe-sense-respond (complex) or sense-analyze-respond (complicated).
Wardley Maps (Wardley)

Strategic mapping visualizes the competitive landscape and component evolution. Maps value chains on an evolution axis (genesis → custom → product → commodity) to identify strategic opportunities, anticipate market changes, and reduce risk by understanding dependencies.

Applied in 4-step process: Step 1 maps current landscape and component positions. Step 2 identifies critical components and dependencies. Step 3 anticipates evolution and identifies strategic moves. Step 4 executes moves based on landscape evolution.

The Multi-Domain Toolkit

Step 1 (General Survey) requires examining problems from multiple perspectives. These are the five domains I draw from:

⚔️
Mission Command / Leadership
West Point / 2/75 Ranger Regiment / Intelligence / Boyd
🚀
Entrepreneurship
Startup Weekend, Lean Startup, VC ecosystem
🏈
Athletics Leadership
Oregon State, Soldiers to Sidelines, coaching
💼
Business Development
VC firms, Drucker, organizational design
🏭
TPS / Shingijutsu Kaizen
Toyota Production System, Operational Excellence

Each domain provides "constituents" that can be broken apart (Step 2) and synthesized across domains (Step 3).

The Process in Action

🎯
Military → Manufacturing

Survey: Examine production bottlenecks from multiple angles.
Condense: Break problem into mission-critical elements.
Synthesize: Apply 2/75 mission planning + intelligence targeting concepts.
Implement: Execute, measure, refine.

📚
Systematic Study

769 books across domains (Drucker, Lean, complexity, AI/ML) provide the "constituent library" for Step 1 (Survey) and Step 3 (Synthesis). Each domain offers pieces that can be recombined.

🤖
LLMs Accelerate the Cycle

LLMs speed up all four steps: faster survey of perspectives, rapid condensation, quicker pattern testing, and accelerated iteration. But the 4-step methodology remains unchanged since 1994.

🏔️

I Like Building Snowmobiles

The Division of Labor Philosophy

🧠
Humans Make Sense

Humans excel at the 3Is: Intelligence (reflection, abstraction), Intentionality (purpose, choice), Identity (how we show up). We navigate complexity, synthesize across domains, and make judgment calls that matter.

⚙️
Machines Compute

LLMs and AI tools accelerate the 4-step cycle: faster survey, rapid condensation, quicker pattern testing, accelerated iteration. They compute at scale but don't replace human sense-making.

The Snowmobile: A human-machine hybrid where the human steers (sense-making, judgment, cross-domain synthesis) and the machine provides power (computation, acceleration, scale). This division of labor is the key to effective AI adoption. Not replacing humans with AI—augmenting human sense-making with computational power.

"Humans are not ants, or birds. We make things, play, and tell stories. Machines help us do it faster."

Multi-Domain Background

Where the Experience Comes From

Diverse experience across military, manufacturing, and continuous learning creates unique problem-solving perspectives.

Military Service (23 Years)

West Point (1992-1995)

Engineering and leadership foundation

2/75 Ranger Regiment (2004-2009)

5 years, multiple combat deployments including Operation Red Wings recovery

Infantry Officer / Military Intelligence (2009-2014)

HUMINT analysis, targeting, counterintelligence, pattern recognition

Key Skills

Systematic analysis, decision-making under uncertainty, structured problem-solving

Business & Operations

Manufacturing Excellence (2022-Present)

Koch Industries: Analyst → Shift Coach → GenAI Champion

Leadership Development (2012-2021)

Oregon State Athletics, Soldiers to Sidelines: coaching, entrepreneurship, veteran transition

Operational Excellence

Lean principles, continuous improvement, cross-domain knowledge transfer

Systematic Documentation

3,000+ pages capturing operational reality, pattern recognition at scale

Academic Background

Arizona State University
MS, Organizational Leadership
University of Pennsylvania
Wharton Executive Education
US Military Academy
West Point (1992-1995)
Case Study

The Strategic Game in Action: Monticello

A 2-year AI-enabled organizational transformation at Georgia-Pacific Monticello paper mill, demonstrating Boyd's 4-step methodology from initial survey to $546M enterprise-scale opportunity.

$64M
Annual Opportunity
Monticello facility validated results
$546M
Enterprise Scale
400 sites, recoverable value/year
Zero Capex
No Pilot Needed
Uses existing Microsoft stack
1

General Survey (Feb-Sep 2023)

Examined paper mill operations from multiple perspectives: manufacturing processes, training systems, leadership development, organizational culture, AI technology landscape, and operator knowledge capture.

Domains Applied:
Mission Command (operator empowerment) + TPS/Kaizen (continuous improvement) + Intelligence (pattern analysis) + Entrepreneurship (pilot approach) + Business Development (stakeholder alignment)
2

Condense to Essential Elements (Oct 2023-Mar 2024)

Broke problem into core components: downtime patterns, operator knowledge gaps, documentation needs, adoption barriers, cultural resistance. Identified key question: "Can we use the Kata?" (Toyota improvement methodology). Condensed to mission-critical insight: Zero adoption in technology-first approach.

Key Finding:
Documentation was "a foreign concept" - cultural challenge, not technology problem. Experience gaps in workforce (33, 29, 26 years then drops). Need operator buy-in first, technology second.
3

Place in Strategic Perspective (Apr-Nov 2024)

Synthesized patterns ACROSS domains to create new approach. Cross-domain synthesis: 2/75 Ranger after-action reviews → Shift knowledge capture system. Mission Command operator empowerment → Bottom-up AI adoption. TPS "respect for people" → Operator memory scaffolding (not replacement). Boyd's OODA loop → Real-time operational learning.

Strategic Pivot:
Shifted from "deploy AI technology" to "embed judgment capture in daily flow." Use existing tools (Microsoft Copilot, OneNote, Teams). Make operators the experts, AI the assistant. Zero capex, operator-led mesh.
4

Implementation (Aug 2024-Apr 2025)

Embedded as Shift Leader to capture real data. Documented 1,408.7 hours of downtime across all shifts. Quantified $64M annual opportunity at Monticello. Scaled to enterprise level: $546M/year potential across 400 sites. Validated with live field data, not simulations. "No Capex. No Pilot. Just Recovered Execution."

Measured Results:
Recovery target: 30 hours/year (2 shifts) = $1.365M/year per facility. Enterprise-level, bottom-up, PBM-aligned system. Not hypothetical—already delivering documented results.

Let's Connect

Interested in discussing cross-domain problem-solving approaches? Want to learn more about the methodology? Let's talk.

Current Focus: Sharing practical approaches to multi-domain problem-solving, building teachable methodologies, and demonstrating how decades of systematic work create unique capabilities.