Transformational Creativity (Boden)

In the AI Story

Boden's canonical examples: the shift from Ptolemaic to Copernican astronomy, the invention of perspective in Renaissance painting, the move from classical to quantum physics. In each case, the new framework did not emerge as a more efficient search within the old one, nor as a combination of existing elements. It required breaking something — the assumption of geocentrism, the convention of flat pictorial space, the demand for determinate particle trajectories — and replacing it with something that made different possibilities visible.

The distinction matters acutely for AI evaluation because current systems demonstrate impressive exploratory and combinational capacities but have not demonstrated transformation. They generate within spaces humans constructed; they connect across spaces humans delineated; they have not, as far as Boden can determine, created a space that did not exist before. Whether they eventually will is among the most consequential open questions in AI research.

The capacity for transformational creativity connects to the candle in the darkness argument in The Orange Pill. Transformation requires dissatisfaction with existing frameworks, and dissatisfaction requires stakes — the kind of caring that a being with interests in the world exhibits. Whether machines can have stakes in this sense is the deepest philosophical question the AI transition raises.

The practical consequence: if exploratory creativity has commoditized and combinational creativity is now abundant through AI partnership, transformational capacity becomes the scarce resource. The humans who retain it — and the institutions that protect its development — will shape the direction of what comes next. The humans who lose it, who optimize their cognition entirely for AI-assisted production within established frameworks, will find themselves irrelevant to the changes that matter most.

Origin

Boden introduced the concept in The Creative Mind (1990) and developed it through subsequent decades, particularly in Creativity and Art (2010). The framework draws on Thomas Kuhn's paradigm shifts but generalizes beyond science to all creative domains, and specifies the cognitive mechanisms involved rather than treating the shifts as mysterious.

Key Ideas

The space itself changes. Not exploration within it, not combination across spaces — the rules, boundaries, and definitions of the domain are rewritten.

Three operations. Recognition that the current framework is inadequate; imagination of a framework that does not exist; evaluation that the new framework is genuinely better.

Requires consciousness. Each operation demands capacities — dissatisfaction, imagination, evaluation of adequacy — that presuppose stakes in the world.

Current AI has not demonstrated it. Impressive exploration and combination, no transformation that Boden recognizes as genuine.

The surviving scarcity. As the other modes commoditize, transformational capacity becomes the resource that determines whose creative work continues to matter.

Debates & Critiques

The most contested question: can AI develop transformational capacity? Optimists point to emergent capabilities in large language models and argue the distinction may dissolve with scale. Skeptics — Boden among them — argue that transformation requires something more than more parameters, namely the kind of stakes in the world that generate genuine dissatisfaction with frameworks. The question remains open, with profound implications for how the AI transition unfolds.