Worldliness and causal binding in intelligent machines: a Heideggerian response to Dreyfus

Large language models challenge Hubert Dreyfus's claim that AI cannot match human know-how. Reinterpreting Dreyfus through Heidegger, the paper argues the key difference is not rules versus skilled coping but correlational patterning versus teleological causation. This teleological causation, called causal-intentional binding, is Dasein's future-oriented act of opening meanings for which it remains ethically responsible. Modern AI achieves contextual fluency but remains stuck in Heidegger's idle talk, lacking the binding that turns coping into conscience. The paper distinguishes two horizons: AI-as-curator, a static mirror organizing human insights, and AI-as-co-evolver, which might acquire proto-teleology through self-referential causal modeling, approaching an artificial Dasein. The consciousness debate thus shifts from embodiment to ontological directionality.

Study at a glance

Abstract

Generative language models seem to refute Hubert Dreyfus’s verdict that AI can never match human “knowing-how.” Re-reading Dreyfus through Heidegger, this paper argues that the decisive divide is not skilled coping versus rules, but correlational patterning versus teleological causation. The latter is termed causal-intentional binding —Dasein’s future-oriented act of opening meanings for which it remains ethically answerable. Derived from the care-structure ( Sorge ) that Heidegger identifies as the being of Dasein, causal-intentional binding names the temporal unity in which thrownness, projection, and falling are gathered into a single directional arc of engagement with the world. While modern AI masters contextual fluency, it remains caught in Heidegger’s idle talk, lacking the binding that turns coping into conscience. Tracing this gap clarifies two horizons: AI-as-curator, a static mirror that organizes human insights, and AI-as-co-evolver, a system that might acquire proto-teleology through self-referential causal modeling, edging toward an “artificial Dasein.” The paper thus relocates the consciousness debate from embodiment to ontological directionality and sketches constitutive and ethical criteria for judging genuine machine worldliness.