The classification is descriptive, not prescriptive. Gardner's argument is that these eight capacities meet the empirical criteria for cognitive autonomy, not that education or policy should develop them in equal measure for all learners. The prescriptive question — which intelligences to cultivate given the AI transformation — is addressed separately, through the five minds framework and this book's educational chapter.
The AI-era diagnostic divides the eight into three groups. Strongly amplified: linguistic and logical-mathematical — the capacities LLMs engage most directly. Partially amplified: spatial and musical, via multimodal systems, though the amplification remains linguistically mediated. Essentially unamplified: bodily-kinesthetic, interpersonal, intrapersonal, naturalistic — the capacities that operate through representational systems the model does not natively inhabit.
This asymmetry has practical consequences. The engineer whose primary cognitive contribution is linguistic-logical experiences AI as profound amplification. The therapist whose primary contribution is interpersonal-intrapersonal experiences it as marginal assistance at best. The craftsperson whose primary contribution is bodily-kinesthetic experiences it as threat or irrelevance. Each experience reflects the structural selectivity of the amplifier, not differential quality of the capacities.
The convergence thesis — developed in this book's final chapter — holds that the most consequential human performances require the integrated deployment of multiple intelligences. The twelve-year-old's question 'What am I for?' arises from the convergence of intrapersonal self-examination, interpersonal need for recognition, linguistic articulation, and existential gravity. No individual intelligence could produce it. This convergence is precisely what the amplifier cannot provide, because the amplifier amplifies parts.
Gardner proposed the original seven in 1983, added naturalistic in 1999, and has consistently resisted further expansion despite pressure to include candidates like existential, moral, and pedagogical intelligences. His position is that the eight meet the criteria clearly; others are under continuing evaluation.
Eight criteria met. Each intelligence satisfies Gardner's empirical threshold for cognitive autonomy.
Three AI-era tiers. Strongly amplified (linguistic, logical-mathematical), partially amplified (spatial, musical), essentially unamplified (bodily-kinesthetic, interpersonal, intrapersonal, naturalistic).
Descriptive not prescriptive. The taxonomy maps cognitive terrain; educational choices are a separate question.
Convergence thesis. Consequential human performances require integration across multiple intelligences — precisely what the amplifier cannot produce.
Resistance to expansion. Gardner has resisted adding further intelligences, preferring clarity about the eight.