Business Pill 21 · Who holds the control
Whoever defines how others connect shapes everything built on top. That is why extensibility is control.
A short explainer video. The numbers in it are round numbers for illustration.
Closed and Extensible Products
Some products do only what their maker built: a fixed set of features. If you need something else, you wait.
Other products can be extended. They offer connection points where anyone can plug in a new tool, a new source of data, a new skill.
The Wall Socket
Think of a wall socket. The maker does not need to know what you plug in. But notice who holds the power. The owner of the socket decides its shape. Everyone else has to build plugs that fit.
Change the socket, and every plug has to change with it. That is why extensibility is control. Whoever defines how others connect shapes everything that gets built on top.
A Worked Example
The video gives an illustration. A closed assistant ships with 10 tools, all built by its maker. An open one ships with 10 and lets others add 1,000 more. That is 1,010 in total, which is this publication’s addition.
The second product grows without its maker doing the work, and its maker still sets the rules.

Why It Matters
This is why the strongest position is often not the best product. It is the product that everything else connects to. Leaving it means rebuilding every connection.
There is one more case. Sometimes the socket is a shared standard that no single company owns. Then a tool built once works everywhere, and the control is spread among everyone who uses it.
Three Questions to Ask
- Where can others plug in?
- Who decides the shape of the socket?
- What happens to us if it changes?
See It in the News
Claude Code Mods: Extensibility and Control Are the Same Layer. A news piece on how an extension point can be the control point.
More Business Pills
- The Memory Wall: Why a Faster AI Chip Is Not Faster AI
- Tokens per Watt: What an AI Data Centre Actually Produces
- Why AI Can’t Be Both Instant and Cheap: Latency vs Throughput
- The Model and the Harness: Why Same-Model Products Differ
- Context, Not Capability: Why a Smart AI Model Gives Poor Answers
- Discardable Software: When Code Is Cheap Enough to Throw Away
- Human in the Loop vs Human on the Loop: Supervising AI Agents
- The AI Audit Problem: When Making Work Is Cheaper Than Checking It
- AI Evaluation as Acceptance Test: How to Know It’s Good Enough









