Eight months ago I rebuilt my own business on a team that does not sleep, does not forget, and does not exist.
I mean that literally. The research, the drafting, the follow-up, the operations that used to eat my week, most of it now runs on an AI team I built and manage. Patient zero was me. Before I told a single client where AI belonged in their business, I had already bet my own on the answer.
Here is what I learned doing it, and it is the opposite of what most people selling AI will tell you.
The part that worked was the part AI never touched.
Automation replaces people. Augmentation multiplies them.
Most of what gets sold as “AI for your business” is automation in a nicer outfit. Find a person doing a task, replace the person. That story is easy to pitch and easy to picture, which is exactly why it keeps failing.
MIT’s 2025 report on enterprise AI, The GenAI Divide, found that about 95 percent of corporate generative AI pilots delivered little to no measurable impact on the business. Ninety-five percent. Not because the models were weak. Because companies bought tools, handed out logins, and walked away. They treated a behavior change like a software install.
Augmentation is the other path. You do not replace the person. You take the load off everything around them so their judgment reaches further than it could alone. The person stays at the center. The AI does the carrying.
That is not a feel-good distinction. BetterUp and Stanford’s Social Media Lab ran a two-year study of 12,000 workers and found the ones with real agency over how they use AI, who treat it as a tool they direct rather than a thing handed to them, were 3.6 times more productive than the ones just given access. Access is not adoption. The multiplier lives in the human, not the software.
The human core is the moat
The reason augmentation works is the reason automation fails. There is a core to your business that AI cannot do, and it is the most valuable thing you own.
In a service business, that core is judgment. The read of the room. The call you make when the client says one thing and means another. The relationship that took years to earn. A client is not paying you for output. They are paying you for the part of the work only you can do.
So the question I ask every leader I work with is not “where can I put AI.” It is “where should AI never go.” Find the human core, draw a hard line around it, and protect it. Then take everything that is not that core, the research, the drafts, the chasing, the admin, and hand it to the machine.
That is the move. AI handles the load. Your judgment handles the call. The core gets more of you, not less.
I have watched what this does in practice. One coaching client went from publishing in fits and starts to showing up everywhere her clients look, because the production system carried the volume and she carried only the thinking. Another felt it in the room, more present with clients, because the work that used to drain the week was no longer hers to do. Neither of them got replaced by AI. Both of them got multiplied by it.
Where I sit now
I help leaders of growing service businesses figure out where AI belongs in their company and where it does not. That is the whole job, and it has three legs.
Strategy, deciding where AI earns its place and where it would do damage. Implementation, building the systems that actually run. And the leg almost everyone skips: people. Getting your team to use this well, so your judgment stays in the loop instead of getting designed out of it.
That third leg is the differentiator, and it is not an accident that it is mine. I spent a career making technology adoptable by actual humans before I ever called any of it AI. Adoption is a people problem. It always was.
This is what davidchung.ai is for. The advisory side, the part where I help you think it through. KyberFive, my company, is the build side that makes it real. You can hire the thinking without the building, or both. Most leaders start with the thinking, because the expensive mistakes happen before anyone writes a line of code.
Start with the blind spots
You do not need an AI strategy deck. You need an honest look at how your business actually runs right now, and where the leverage is hiding.
That is what the diagnostic is. Thirty minutes, no pitch. I look at how your business operates and tell you where AI would multiply you and, just as important, where it would quietly wreck the thing your clients are paying for. AI is very good at finding the blind spots in how you are running your business. Usually they are not where you think.
If you want that look, book the diagnostic on this page. Bring the one workflow that eats your week. We will start there.
Frequently Asked Questions
What is the difference between AI automation and AI augmentation?
Automation replaces a person at a task. Augmentation multiplies a person by handling the parts of the work that do not need their judgment, so their judgment reaches further. The distinction matters because most failed AI projects tried to automate a job that still needed a human in the loop. Augmentation keeps the human at the center and gives them leverage, which is why it tends to stick where automation stalls.
What does a fractional Chief AI Officer actually do?
A fractional CAIO owns the strategy, implementation, and adoption of AI inside a business without being a full-time hire. The work has three legs: deciding where AI belongs and where it does not, building the systems that put it to work, and getting the people to actually use it well. The third leg is the one most companies skip, and it is where most of the value and most of the failure lives.
Why do most enterprise AI pilots fail?
MIT’s 2025 report The GenAI Divide found that roughly 95 percent of enterprise generative AI pilots delivered little to no measurable impact on the business. The common thread is not bad technology. It is that companies bought tools and handed out logins without teaching anyone how to fold AI into how they actually work. They treated a behavior change like a software install, and skipped the adoption work entirely.
Can AI replace human judgment in a service business?
No, and trying to is the mistake. In a service business the human core, the judgment, the relationship, the read of the room, is the thing the client is paying for. AI can take the load off everything around that core so the core gets more of your attention, not less. The goal is to protect the part only a human can do and automate the rest.
Where would AI multiply you, and where should it never go? That is the question I help leaders answer. Start with the 30-minute diagnostic on this page.