Systems Thinking and Strategic AI Deployment
Former Marine Corps officer David Wood reveals why 90% of AI projects stall between pilot and production—and it's not a technology problem. Drawing from 45 years of systems engineering, Wood shares military frameworks for AI deployment: the production readiness test, triage methodology for investment decisions, and governance as rules of engagement. Learn when AI should have authority versus advise, and the one metric that actually predicts enterprise AI success.
Paul Brzozowski
Founding Team Member • Olakai
Paul Brzozowski is a seasoned technology executive and founding team member at Olakai, where he drives strategic innovation in product development and business growth. With extensive experience in emerging technologies and entrepreneurial ventures, Paul brings deep expertise in scaling startups and developing transformative digital solutions. His background spans multiple technology sectors, with a proven track record of building high-performance teams and delivering cutting-edge technological products.
LinkedIn ProfileDavid Wood
Former Marine Corps Officer, Systems Engineering Expert • Gladwood LLC
David Wood is a seasoned systems engineering professional with a distinguished background as a former Marine Corps Officer, bringing rigorous strategic leadership and technical expertise to complex technological challenges. As the founder of Gladwood LLC, he leverages his military and engineering experience to help organizations optimize their technological systems and operational strategies. With a proven track record of translating technical complexity into actionable insights, David is a respected consultant who bridges the gap between advanced engineering principles and practical business implementation.
LinkedIn ProfileMost AI projects fail. Not because the technology isn't ready - but because organizations treat AI like a tool instead of a system. David Wood brings 45 years of systems thinking to the enterprise AI conversation, and his perspective cuts through the noise: AI is a technology, but AI deployment is a human challenge.
Former Marine Corps officer with a master's in systems engineering. 25 years selling enterprise technology. Now running Gladwood LLC, helping organizations understand human performance using the Hartman Value Profile while advising AI companies like Neuroscale AI on strategy.
When you've spent decades studying what makes complex systems succeed or fail, the answer is always the same: the human factor.
In This Episode
The Three Reasons AI Stalls
- No clear mission and metric (trying AI vs. cutting cycle time by 30%)
- Not wired into the system (stuck in the swivel chair gap)
- Human trust issues (people nod and keep doing things the old way)
The Production Readiness Test from the Marines
- Can the operator trust it under pressure?
- Is it integrated with everything else in the stack?
- Who owns it when it fails at 2 AM?
The One Metric That Actually Matters
- Time to a mission-critical outcome at equal or better quality
- Why this rolls up speed, quality, cost, and competitive advantage
- How to move from "usage metrics" to real business impact
Military Triage Applied to AI Investment
- Treat Now: Mission-critical with 60-90 day impact
- Stabilize: Good alignment, something missing
- Observe: Cheap experiments at the edge
- Let Go: Misaligned pet projects
- The filter: Mission and readiness, not hype and novelty
Governance as Rules of Engagement
- Why governance isn't bureaucracy - it's protection
- Clear mission and boundaries for AI authority
- Ownership and accountability (AI is the junior Marine, never the commander)
- Feedback and adaptation as a habit, not a document
Three Military Frameworks for Enterprise AI
- OODA Loop: Where should AI live in your observe-orient-decide-act cycle?
- Commander's Intent: AI needs a mission, not just a model
- After Action Reviews: Build learning into your operating model
When AI Should Have Authority vs. Advise
- Four questions: Risk? Reversibility? Rule clarity? Accountability?
- The principle: AI can act, but it can never be accountable
- Only people and organizations can be accountable
Key Quotes
"Most AI work never escapes the science project phase. You get a great demo, maybe a clever pilot, a couple of excited champions, and then nothing really changes."
"Governance as rules of engagement weren't there to slow us down. They were there to make sure our power was applied the right way, at the right time, on the right target."
"The big leap in the next 12 months is from cute assistant to reliable co-worker."
Timestamps
- 00:00 - Introduction: 45 Years Studying Complex Systems
- 03:42 - From Marine Corps to Enterprise Technology
- 08:15 - The Systems Engineering Lens on AI Deployment
- 14:20 - Core Question 1: Why AI Stalls Between Pilot and Production
- 19:45 - Core Question 2: The One Metric That Matters
- 25:10 - Core Question 3: Military Triage for AI Investment
- 31:35 - Core Question 4: Governance as Rules of Engagement
- 37:50 - Core Question 5: From Cute Assistant to Reliable Co-Worker
- 44:15 - Deep Dive: Three Military Decision-Making Frameworks
- 52:30 - Deep Dive: When Should AI Have Authority vs. Advise?
- 58:40 - Closing: The Systems Challenge Requiring Military-Grade Discipline
About the Guest
David Wood is a former Marine Corps officer with a master's in systems engineering from Naval Postgraduate School. He spent 25 years in enterprise technology sales and business development. Now he runs Gladwood LLC, helping organizations understand human performance using the Hartman Value Profile, and advises AI companies like Neuroscale AI on strategy. His perspective: AI deployment requires military-grade discipline, systems engineering rigor, and deep understanding of the human factor.
About Enterprise AI Unlocked
Enterprise AI Unlocked explores how organizations move from AI experimentation to measurable business impact. Hosted by Paul Brzozowski from the founding team at Olakai, each episode uses a consistent five-question framework to unpack what actually works when deploying AI at scale. New episodes weekly.
