Human First. AI Second. (Always.)
Why AI integration keeps failing quietly—and how design thinking may finally fix it. Learnings from a recent AI-human integrations work session.
Last year, I was obsessed with AI.
If you didn’t notice… congrats, you probably weren’t on Zoom with me at 2:00 a.m. while I whispered lovingly to yet another tool.
Tool #37.
Pure FOMO.
“This one will save me.”
Spoiler: it did not. It mostly stole my sleep and made me fluent in release notes, which is not a personality anyone asked for.
I am not a tool builder.
I am a brand builder.
And that realization? It wasn’t failure. It was clarity.
“I didn’t fall in love with AI tools. I fell in love with what they revealed about humans at work.”
The real turning point
The switch flipped when I ran my first AI discovery and integration workshop. It wasn’t about prompts and plugins. It was about people trying to figure out where they still mattered.
Facilitation has always been my superpower, and suddenly it was obvious: AI doesn’t break workflows first. It breaks people systems first.
My 30 years of brand building, behavior decoding, and design thinking facilitation finally lined up with this moment. AI wasn’t asking for more tech. It was begging for better design.
Why most AI rollouts quietly die
Most AI programs don’t crash and burn.
They slowly fade into “remember when we tried that?” territory.
A pilot launches. A few teams play. Leaders feel proud. Productivity bumps in pockets, confusion spreads everywhere else.
Six months later? The tools are still there, but the transformation is not.
“AI reassigns responsibility without ever renegotiating authority—and humans feel that instantly.”
So people protect themselves. They fake adoption. They nod in meetings, then go back to doing things the old way in a secret Google Drive folder.
This is clearly resistance to badly designed change.
Why design thinking is the antidote
Design thinking gets framed as sticky-note theater, but at its core it’s a way to help humans make sense of chaos. AI is chaos with a user interface.
It blurs roles. Compresses decisions. Shifts expectations overnight. Design thinking slows things down just enough for people to catch up emotionally, not just operationally.
“Used properly, design thinking isn’t brainstorm theater. It’s the operating system for AI adoption.”
The job is not to brainstorm more tools. It’s to redesign how humans and AI work together without breaking trust.
Five design moves I have test driven that make AI actually stick
1. Start with human risk, not use cases.
Skip “what can this tool do?” and start with “what do people fear losing?” Relevance, control, credibility, decision power. Until those risks are named, no roadmap will land.
2. Map decision flow before workflow.
If teams can’t answer “who decides what, and when?” AI will happily guess. That’s how chaos starts—not because the model is wrong, but because authority never was.
“If you don’t design who decides, AI will—and no one will like its answers.”
3. Redefine what ‘good work’ means.
AI breaks your old status signals. Speed isn’t effort. Output isn’t thinking. Busy isn’t a contribution. If leaders don’t redefine excellence, people will cling to old behaviors while feeling behind anyway.
4. Prototype trust, not tools.
Early AI experiments shouldn’t just ask “does this work?” They should ask:
When do we rely on AI?
When do we override it?
Who owns the outcome when it’s wrong?
Trust is not a feeling. It is a system.
5. Design the learning loop.
AI integration isn’t a launch; it’s a feedback system. Models evolve, context shifts, teams learn. A designed learning loop keeps the organization adaptable instead of brittle.
This is brand work, not IT work
How a company integrates AI signals what it values. Employees notice first. Customers feel it eventually.
“AI adoption becomes culture—whether you design it or not.”
When you apply design thinking to AI integration, something important shifts: people aren’t scared to engage. They’re willing to try, to fail safely, to see where they fit in the future.
That’s when adoption actually happens.
Who knew one day the real work wouldn’t be “using AI better,” but building the conditions where humans can use AI while staying human.
If you want a sneak peek at the work session where all of this turns into exercises, questions, and actual decisions, shoot me a note.