Genuine vs. Mediated Competence
Genuine competence is built through the effortful process of learning, failing, adjusting, and learning again. It is embodied — deposited in the neural pathways, the procedural memory, the intuitive responses of the individual who has earned it through practice. Mediated competence, by contrast, is borrowed capability — the ability to produce results that depend on an external tool. The child who can direct an AI assistant to produce high-quality work has a real skill that has value in the world as it is. But she has not had the developmental experience that produces the felt sense of industry Erikson's framework requires. Direction is not the same as making. Management is not the same as mastery. The distinction is invisible in the product and decisive in the self.
In the AI Story
Two essays, one written by a child through hours of struggle and one generated by a machine in seconds, may be indistinguishable to an outside reader. The words on the page are not the developmental variable. The experience of producing them is. The child who has struggled has developed a relationship to writing — a feel for language, a sense of rhythm, an awareness of how words work together — that is not visible in the product but is constitutive of her developing self.
The distinction matters most where the cogwheel propagates. The child who emerges from the Industry stage with genuine competence enters the Identity stage from a position of psychological strength — she knows what she is good at, what she cares about, and what she has to contribute. The child who emerges with only mediated competence enters from a position of hidden vulnerability. She may appear confident because her outputs are impressive, but the confidence rests on the continued availability of the tool that produces them.
The distinction is not a rejection of AI. It is an insistence that developmental environments must preserve opportunities for genuine productive struggle alongside AI-augmented learning. The child must first write, badly and with difficulty, before she can learn to evaluate writing. She must first draw, clumsily and with frustration, before she can learn to see what makes a drawing work. Production precedes evaluation in the developmental sequence, and the sequence cannot be reversed without cost.
Erikson's framework does not require the elimination of AI from childhood. It requires the recognition that the tool's benefits at the level of output are purchased with developmental costs at the level of the self — costs that are invisible in the short term, decisive in the long term, and reversible only through the deliberate design of environments that make genuine struggle possible.
Origin
The concept is a 2025–2026 extension of Erikson's framework, grounded in his distinction between external capability and internal competence. It responds to the empirical observation that AI-generated output has reached quality levels where human evaluators cannot distinguish between the product of machine and human production without disclosure.
Related concepts include Shannon Vallor's moral deskilling, Anders Ericsson's distinction between naive and deliberate practice, and the cognitive-science literature on desirable difficulties.
Key Ideas
The product is not the developmental variable. Two indistinguishable outputs can represent categorically different developmental experiences.
Genuine competence is embodied. It lives in procedural memory and intuitive response, not in the ability to produce outputs.
Mediated competence is borrowed. It depends on the continued availability of the tool; it does not transfer when the tool is absent.
The cogwheel propagates the difference. The adolescent whose childhood competence was mediated enters identity formation from a position of hidden vulnerability.
Production must precede evaluation. The sequence cannot be reversed; judgment requires experiential foundation.
Appears in the Orange Pill Cycle
Further reading
- Shannon Vallor, Technology and the Virtues (Oxford University Press, 2016)
- Anders Ericsson and Robert Pool, Peak (Houghton Mifflin Harcourt, 2016)
- Robert Bjork and Elizabeth Bjork, 'Making Things Hard on Yourself, But in a Good Way' (2011)
- Nicholas Carr, The Glass Cage: Automation and Us (W.W. Norton, 2014)