Independence Ratio

In the AI Story

The ratio is empirically simple but practically uncomfortable. Measuring it requires removing the scaffold and observing what the builder can do alone. And removing the scaffold, in the current environment, feels — as Segal confesses in his epilogue — like 'voluntarily diminishing yourself.' The discomfort is diagnostic: if the removal feels like loss rather than test, the scaffold has been functioning as prosthesis.

Three complementary measurements operationalize the ratio. First, independent capability over time: the trajectory of what the builder accomplishes without AI, measured at regular intervals across months and years. Second, transfer to novel problems: the builder's ability to apply understanding constructed during AI-augmented work to problems the AI has not encountered. Third, metacognitive awareness: the builder's ability to distinguish between what she knows independently and what she knows only through the scaffold.

The structural problem the ratio reveals is that no one has incentive to measure it. AI companies are disincentivized — a measurement that might show their tools produce prosthesis rather than scaffolding threatens the business. Users are disincentivized — testing independent capability might reveal uncomfortable gaps between perceived and actual competence. Organizations are disincentivized — managers reluctant to ask teams to spend a week without AI for the sake of assessing unaugmented performance.

The ratio is also the only measurement that distinguishes scaffolding from prosthesis. Both produce impressive supported performance. Both look identical from outside. Only the withdrawal test — only the ratio of independent to augmented capability and how that ratio changes over time — reveals whether months of partnership built something internal or merely something external.

Origin

The concept is explicitly formulated in the Bruner volume, drawing together Bruner's lifelong insistence that the measure of effective scaffolding is independent performance after withdrawal. Related empirical work on AI cognitive effects includes the MIT Media Lab's 2025 cognitive debt research and earlier deskilling literature in aviation and medicine.

Key Ideas

Simple ratio. Augmented capability divided by unaugmented capability, measured over time.

Direction matters. Rising independence indicates scaffolding; stagnant independence with rising augmentation indicates prosthesis.

Three measurements. Independent capability trajectory, transfer to novel problems, metacognitive awareness.

Structural disincentives. Commercial, personal, and organizational forces all discourage measuring the ratio.

Withdrawal as test. Only removing the scaffold reveals whether internalization has occurred.

Debates & Critiques

Whether the independence ratio is a meaningful metric in a world where AI is effectively always available is contested. Some argue that unaugmented capability is increasingly irrelevant — like asking how fast a commuter can walk to work when everyone drives. Bruner-aligned thinkers respond that 'always available' is a current condition, not a permanent fact: tools fail, problems exceed training data, and the builder who cannot operate without the scaffold is structurally fragile regardless of how available the scaffold usually is.