AI Exposes Leaders; The Cultivation Test
+Give your team AI tools and watch what happens. What you see is who you already were.
+ +AI does not create leadership. It exposes it.
+ +This is the uncomfortable truth behind every enterprise AI rollout. The organizations that cultivated their people before the tools arrived are now watching their teams multiply in impact. The organizations that stripped them for short-term yield are watching the same tools accelerate a collapse they had already started.
+ +AI is a cultivation test. And most leaders are failing it without even knowing they took it.
+ +The two experiments running in parallel right now
+ +Picture two engineering teams. Both receive access to the same coding assistant, the same documentation tool, the same automation platform.
+ +Team A has a culture of trust, context-sharing, and psychological safety. Their senior engineers already spent time teaching juniors how to think about tradeoffs. When the AI tool arrives, the seniors use it to remove boilerplate so they can focus on architecture. The juniors use it to learn faster because they have mentors who help them interpret what the model generates. The team's output increases; their quality increases; their engagement increases.
+ +Team B has a culture of output metrics, individual accountability, and implicit competition. Their senior engineers guard the interesting work. When the AI tool arrives, management sees an opportunity to reduce headcount. The seniors are told to "just review AI output." The juniors are given the model and expected to produce without support. Rework increases. Escalations increase. The seniors who remain are now editors of code they did not write, fixing mistakes they would never have made. Attrition follows.
+ +Same tools. Opposite outcomes. The variable was never the AI.
+ +What the cultivation test reveals
+ +The cultivation test is simple. Give your team powerful tools and observe:
+ +-
+
- Do they use them to amplify judgment, or to bypass it? +
- Do senior people become teachers, or do they become bottlenecks? +
- Does trust increase, or does surveillance increase? +
- Do your best people stay, or do they leave? +
The answers are not about the technology. They are about the culture that existed before the vendor showed up.
+ +Why this matters for your positioning
+ +If you are a leader right now, you are being tested. Not by a board presentation; not by a roadmap review. You are being tested by what your team does with the tools you just gave them.
+ +If you cultivated them, AI is the multiplier you always wanted. Your institutional knowledge stays in the building. Your people get faster without getting shallow. Your senior engineers become force multipliers instead of human single points of failure.
+ +If you strip-mined them, AI is an accelerant on a fire you started. The knowledge walked out months ago. The model is now generating answers no one in the room can verify. Your best people are interviewing elsewhere because they can see where this ends.
+ +This is the core argument of both my books
+ +In The Soil Beneath the Code, I write about the infrastructure of trust and context that makes engineering organizations durable. That soil is what determines whether AI helps you or hollows you.
+ +In A Walk in Mentoring, I write about the specific practice of mentorship as the highest-leverage way to build that soil. It is not a soft skill. It is a competitive advantage that compounds faster than any technology you can buy.
+ +Together, they frame the choice every leader is making right now; not whether to adopt AI, but whether your organization is the kind of place where AI adoption actually works.
+ +The challenge
+ +Here is the test. Do not wait for a consultant to tell you the answer. Look at your own team this week.
+ +Watch what happens when a junior engineer uses a coding assistant. Do they have someone to ask "why did it suggest this?" Or are they flying blind?
+ +Watch what happens when a senior engineer saves three hours with automation. Do they invest those hours in teaching? Or do they just pick up the next ticket?
+ +Watch what happens when a model produces a plausible but wrong answer. Does someone in the room know enough to catch it? Or is the room already empty of that knowledge?
+ +What you see is who you already were.
+ +The good news: cultivation is a choice you can still make. The soil is not dead until you stop tending it.
+ +The bad news: every quarter you keep stripping, the harder the replanting gets.
+ +AI exposes leaders. The question is not what tool you bought. The question is what kind of leader the tool just revealed.
+