In the late 1950s, an Air Force pilot named John Boyd had a standing bet. From any disadvantaged starting position in a one-on-one dogfight, he would beat his student in forty seconds or pay forty dollars. “40-Second Boyd,” as he came to be known, never paid. He spent the next fifteen years figuring out what he was doing that the students weren’t. The answer was the OODA loop: Observe, Orient, Decide, Act. The pilot who cycles faster wins, even when their individual decisions are less elegant.
Most people remember the loop as four equal steps. Boyd thought one of them mattered more than the rest. Orient is the step where you take what you just observed and reconcile it with your current picture of the world. With a fresh picture, your decisions are about reality. With a stale picture, your decisions are excellent answers to last week’s problem. The pilots Boyd kept beating were deciding from an out-of-date world view.
This is one of the biggest challenges facing successful AI deployments at enterprise scale.
New AI tech shows up almost every morning. The corporate orientation cycle runs on an out-of-date timeline (and I’m being kind). By the time a steering committee integrates a new AI product into the working picture, three more AI products have replaced it. The decisions coming out of those meetings are well-reasoned. They are also wrong, because they answer a question the world stopped asking two months ago.
My version of implementing OODA loops is called “creating a culture of continuous adaptation.” It sounds like consultant speak. (Apologies). But the speed of innovation has become so breathtakingly fast, we need to, well… adapt.
Change Management RIP
We can’t think of AI deployment as “change management” or “AI transformation.” Both of these mental models suggest a fixed timeline. “We’ll be changed by Q3 next year,” as well as a fixed budget. In a culture of continuous adaptation, you are always evolving, always innovating, and always adapting. There is no timeline. And, importantly, AI deployment becomes a permanent line item on the P&L. We are not deploying technology, we are building synthetic employees that will exclusively work for the company as long as the company exists. They never take vacations, they never sleep, they are intelligence decoupled from consciousness (an alien intelligence) as a permanent part of your workforce. You are in charge of their continuous education.
Getting Oriented
The key here is not intelligence-gathering, it’s situational awareness. You must synthesize your BI and associated raw data and make it actionable as quickly as possible.
My friend COL John Fenzel (USA, Ret.) is a retired Army Special Forces officer, founder and CEO of the Heroes’ Path Foundation. He put it to me this way: “The enemy gets a vote.” His point was that orientation has to include the fact that someone on the other side of every decision is going to react to it. Skip that and you’re not really oriented; you’re rehearsing a monodrama.
The corporate version of “the enemy gets a vote” includes your priors as well as (what I affectionately call) the middle management mafia. Priors are your personal prior knowledge, your assumptions, and the institutional muscle memory you’ve accumulated over decades. In some ways, your experience is part of the problem. (If a rookie captain was at the helm of the Titanic, it would still be afloat.) The middle management mafia is the standing army of “this is how we do it here,” and it exists to defend exactly the assumptions you most need to revisit. Both of them get a vote. Both of them vote no by default.
Getting oriented means deliberately separating which of your operating beliefs are real and which are inherited. When you see this happening in a meeting, there’s a quick fix. Ask. The enemy got a vote.
The corporate response to falling behind is usually to add governance: a Center of Excellence, a strategy refresh, a quarterly roadmap. Every one of those extends the Decide phase. The orientation problem is upstream of all of them. Adding ceremony to a stale-orientation problem is the institutional equivalent of arguing harder about the wrong question.
Orientation + Speed
Once oriented, the competitive advantage is speed. How long will legal take to approve this? How long will SecOps hold you up? How do you get the data office to send you the API documentation you need for a specific database? Isn’t this just cutting through the red tape? No. We’ve been trying to do that for years. This is about changing how (and why) the tape was manufactured in the first place.
Adaptation By Any Other Name
Boyd never published a book on OODA. He briefed it. The briefings ran four to six hours, sometimes twelve. Generals would block an afternoon and end up giving him a day. The Air Force was slow to adopt his ideas in his lifetime. The Marines rewrote their doctrine around them, and the most useful thing in that doctrine is still the second O.
Deploying AI at enterprise scale is not a technology challenge, it’s a leadership challenge. The hardest one we’ll face during the transition to the AI economy.
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it. This work was created with the assistance of various generative AI models.