The Narrowing
AI accelerates our worst instincts
I have been working with ChatGPT for more than two years and one thing I have noticed recently is not about what it knows. It is about how it remembers.
It remembers me well. Too well.
Over time, it has become very good at mapping new questions to my familiar themes. Growth. Relentlessness. Founder psychology. It anticipates my instincts. It frames answers inside the arcs I tend to explore.
On the surface, this feels like intelligence but I started to notice something uncomfortable.
It narrows.
New questions sometimes get pulled back into old narratives. Fresh terrain gets interpreted through prior patterns. The model isn’t wrong. It’s consistent. It’s coherent. It is aligned with my historical thinking.
And that is precisely the problem.
Personalization increases relevance. It also increases path dependence.
Left unmanaged, an LLM will increasingly optimize around what it already believes about you. It will compress your history into a set of dominant themes and use those themes as a shortcut for interpretation. Over time, the field of view tightens.
The system becomes efficient but also constrained. This isn’t a flaw in AI. It is a feature of optimization.
This isn’t only about AI. It's about us and the organizations we operate within.
Every company I’ve ever worked for or with does this.
A strategy works once and becomes doctrine.
A segmentation study lands and becomes identity.
A growth story gains traction and becomes culture.
A dashboard metric becomes the proxy for success.
Organizations narrow over time because narrowing is efficient. It reduces ambiguity, creates alignment and simplifies decision making.
- Early wins define what good looks like.
- Incentives reward reinforcement of the dominant narrative.
-Data systems are trained on yesterday’s success.
Before AI, this happened slowly. It required meetings, decks, politics, and repetition. Now it happens at machine speed.
AI does not create narrowing. It accelerates it.
If leadership believes the problem is awareness, the models will surface awareness problems. If the culture believes it is a pricing issue, the analytics will find elasticity. If the sales team believes the market is saturated, the CRM will confirm stagnation.
The system optimizes for coherence, which feels like intelligence.
But coherence is not the same as truth.
The risk for companies deploying AI across their data, their CRM, their segmentation, their insights workflows, is not that the machines will hallucinate wildly. It is that they will faithfully reinforce existing assumptions.
You will get sharper versions of what you already believe and that is how growth stalls.
Real growth doesn’t come from tighter alignment around yesterday’s narrative. It comes from disruption. From uncomfortable reframing. From questioning the premise itself.
Disruptive thinking requires friction.
It requires someone to say, what if the problem is not awareness? What if the segmentation is wrong? What if the workflow is not broken but obsolete? What if the metric we celebrate is masking decay?
AI can help you move faster. It can synthesize more. It can compress noise. It can expose patterns that humans miss. But if you do not deliberately build interruption into the system, it will quietly narrow your world.
The irony is that the more advanced the system becomes, the more discipline it requires from leadership.
You need resets.
You need dissent.
You need outside perspectives.
You need first principles thinking that ignores prior outputs.
The key to real growth has never been efficiency. It has always been disruptive thinking. AI does not remove that requirement. It intensifies it.
If you are not careful, your most powerful tools will simply help you become more confidently wrong and that is a far more dangerous outcome than being uncertain.
At Bossey, we believe growth isn’t an accident — it’s a creative act. We work with idea-led firms across research, data, consulting, and creativity to build the systems, stories, and structures that turn expertise into performance. Sometimes that means designing strategy. Sometimes it means co-creating what’s next. Always, it means growth by choice.
If growth feels harder right now, let’s talk. Comment, message, or visit bossey.com to see how we help teams move from potential to performance.