LLMs Are Infinite Midwits

Large language models can aggregate and reproduce vast objective knowledge reliably, yet they systematically lack the subjective intelligence that underwrites judgement, moral reasoning, and life‑shaping decisions. As a result, their fluency can mislead users into overestimating their ability to make normative or context‑sensitive calls. — If accepted, this framing warns policymakers, educators, and platform designers to distinguish performance metrics from real‑world judgment and to avoid treating LLM outputs as substitutes for human discretion.