Everyone is talking about AI strategy. Very few are actually using it.
I've seen this movie before. During the print-to-digital shift at Yellow Pages, we had the same meetings, the same fears and the same strategy decks explaining why we didn't need to change just yet.
Different decade. Same behaviour.
Here's the uncomfortable truth that applied then and applies now: Your AI initiative isn't failing because the technology isn't ready. It's failing because the organization isn't.
A few observations from actually being in it:
1. "AI strategy" is often fear dressed up as rigor
Roadmaps. Steering committees. Governance frameworks. Most of this exists to delay the moment when someone has to try something and risk being wrong.
Here's a simple test: If your AI plan can't start this week, it's not a plan—it's avoidance.
Not a rollout. Not a transformation. A real experiment that surfaces friction, feedback and learning.
👉 The organizations "preparing" for AI are mostly preparing not to fail publicly.
2. AI renders legacy workflows insolvent
Teams resist AI tools that can do the work in seconds not because of quality—but because they spent years perfecting the manual way.
This is the sunk cost fallacy, applied to process.
In investing, when a thesis is broken, you sell. You don't hold because you liked the stock in 2015. In business, we cling to "the process" because we paid so much to build it.
AI doesn't politely optimize those workflows. It declares them insolvent.
👉 You're not protecting quality—you're protecting past investment in inefficiency.
3. AI doesn't expose skill gaps—it exposes thinking gaps
When people say, "The output wasn't very good," what they're usually discovering is:
- the problem wasn't clearly defined
- the assumptions were implicit, not explicit
- the work was never truly structured
AI doesn't compensate for vague thinking. It punishes it immediately.
The biggest gains come from forcing clarity—turning fuzzy intent into explicit instruction.
👉 If AI makes your work look messy, it's because it always was.
4. AI threatens roles long before it threatens jobs
AI doesn't just speed things up. It exposes how much work is coordination, not judgment.
When teams start using AI to draft briefs, synthesize options or pressure-test decisions before meetings, something uncomfortable happens:
Meetings get shorter. Pre-reads get sharper. And some roles stop being as necessary as they used to be.
Not because AI replaced them—but because ambiguity no longer protects them.
👉 Resistance to AI is often resistance to accountability.
5. The tool is just a ticker symbol
Companies waiting for the "perfect, secure enterprise AI tool" are missing the point.
The tool is volatile. It will change. What compounds is the capability—the ability to deconstruct work, structure thinking and redesign workflows.
That's your principal capital—not the software.
👉 Obsessing over the tools is how organizations avoid deciding what friction they're actually willing to remove.
Final thought
We've seen this pattern before. Search. Mobile. Cloud.
Legacy companies overthought. Startups moved fast and learned in public.
The winners weren't smarter. They were less afraid of being wrong early.
AI won't reward vision decks, credentials or confidence. It will reward the organizations willing to expose weak thinking, broken systems and fragile egos—first.