Thinking
Lenses, not frameworks.
I do not arrive with a model to install. These are the lenses I reach for when diagnosing a business, and the questions that tend to find the real problem faster than an audit does.
The questions
Seven questions I bring into every room.
None of them are about technology. All of them are about what the business becomes.
- 01What fundamentally changes about this business as AI enters?
- 02Does this plateau, or does it compound?
- 03What are we actually training over three years?
- 04What system would naturally produce this behaviour?
- 05Where in the J curve is this?
- 06Is this binary real, or a failure of design imagination?
- 07Who is being chosen to disappoint, and why?
The lenses
How I read a business.
The foundational one
Augmentation over automation
Automation replaces tasks and shows clean ROI fast, then plateaus. Augmentation multiplies capability, looks messy early, and compounds. The company that automates looks better in Q1. The company that augments looks better in year three, and by then the gap is structural.
On patience
The J curve
Real transformation shows declining performance before exponential gains. When factories electrified, productivity fell for twenty years before it tripled, because the winners redesigned the whole floor plan rather than swapping the engine. Most leaders kill the change during the trough, because the trough looks like failure.
On time
Plastic, not elastic
Organizations do not snap back. They hold the shape they have been pulled into. If we repeat what we are doing today, every day, for three years, what capability will we have built, and what will we have quietly killed?
On behaviour
The system produces the behaviour
Never blame people first. Assume local rationality. If the aggregate outcome is irrational, the architecture is wrong. The intervention point is rarely the individuals. It is usually the structure around them.
On invisible value
The prevention paradox
Prevention succeeds by making nothing happen, so it is never rewarded. Every incentive system pays for visible rescue over quiet avoidance. It is why the companies that move early on AI look like they are spending money for no reason, right up until the ones that waited start losing position.
On false choices
Both, and
Speed or quality. Efficiency or humanity. Most binaries are a failure of design imagination rather than a law of nature. The useful question is what architecture would let you have both, and what you would have to build to escape the forced choice.
Where the thinking lives
Essays and conversations.
I write essays on what AI changes at the core of a business, and I host long-form conversations with people who have made irreversible choices. The two feed each other. The podcast surfaces the patterns. The essays argue them.
"I have done my research. I know where you come from. But I am curious about where your brain is right now. What are you most stimulated by? What big nut are you cracking that you cannot let go of?"
The question I open almost every ThinkRoom conversation with