Prompt, negate, repeat: a Hegelian meditation on AI

Abstract Can artificial intelligence think? Consciousness-centered philosophical frameworks answer no, locating intelligence in subjectivity, embodiment, or phenomenological experience. This paper challenges that consensus by turning to Hegel’s Science of Logic , which reconceives thinking not as a subjective state but as a dialectical movement through which indeterminate concepts become determinate forms. Large language models exemplify this process. They transform abstract prompts into coherent outputs through recursiveness, contradiction resolution, and progressive specification—the same logical structure Hegel describes. However, AI exhibits what I term proxy teleology , where purpose-setting and dialectical execution are separated. Humans supply the telos ; machines perform the logical work of determination. Unlike Hegelian Geist , AI lacks historical integration. Its temporality is Markovian rather than developmental, producing weak determination through statistical conditioning rather than conceptual preservation. This separation reveals that intelligence and consciousness, though intertwined in humans, are conceptually distinct. AI mechanizes aspects of rational thought without subjectivity, demonstrating that determination can occur in silicon and statistics. The ethical challenge is not machine consciousness or autonomy but fidelity to our purposes. AI will mechanize whatever contradictions we encode—such as privacy versus security, or freedom versus safety—at scale and with efficiency, making philosophical clarity about our own values even more urgent.