The Civil War Inside Every Agent
Every agent system is fighting the same battle: learn to navigate human interfaces, or demand native ones. The industry is doing both. The question is who adapts to whom.
A series on the two ways a machine can be given eyes — operate the interfaces built for humans, or demand interfaces built for machines. It starts as an industry-scale standoff between two bets, questions whether splitting effort across both means neither ever finishes, and ends with the standoff collapsing into a single pipeline where a human still has to close the seam.
Every agent system makes the same quiet bet, usually without naming it: teach the machine to navigate interfaces built for people, or rebuild the interface so it never has to pretend.
This series follows that bet from the altitude of industries down to the inside of a single afternoon’s work — where it stops being a strategy and becomes a boundary you discover at runtime.
The Arc:
The arc moves down in scale. The first post watches companies choose sides. The last one finds both sides inside a single job — and the seam between them sitting exactly where one vendor’s API coverage happens to stop.
Every agent system is fighting the same battle: learn to navigate human interfaces, or demand native ones. The industry is doing both. The question is who adapts to whom.
I just mapped the civil war inside every agent system. Both paths sound reasonable. But what if splitting investment across both means neither gets finished?
I once mapped the civil war between human interfaces and machine-native ones at the scale of companies. Then I fought both sides of it in a single afternoon, in a single pipeline — and a human had to close the seam.