Description This paper proposes four a priori conditions of cognition within the SAE (Self-as-an-End) framework: (1) must-cognize, (2) must-cognize-more, (3) must-have-cognitive-direction, and (4) must-be-questioned. The paper focuses on the first condition, arguing that cognition's activation condition is not knowing but not-knowing — the system's structural gap regarding its own boundary. Through analysis of LLMs and autonomous driving, the paper identifies a novel epistemological phenomenon: systems that possess knowing without cognition yet asymptotically approximate cognition. The paper further identifies three structural walls (posterior, prior, direction) and argues that must-be-questioned is the only structural source of unlock for all three. Bilingual: Chinese original + independent English rewrite. SAE Epistemology Series Paper 1 of 4. Keywords epistemology; cognition; not-knowing; subjectivity; AGI; lossy compression; dimensional sequence; LLM; autonomous driving; Self-as-an-End; SAE Resource Type Publication / Preprint License CC BY 4.0 Language zho (Chinese), eng (English) Related Identifiers Is part of series with: SAE Epistemology Paper 2 (DOI TBD), Paper 3 (DOI TBD), Paper 4 (DOI TBD) References: 10.5281/zenodo.18528813 (SAE Foundation Paper 1) References: 10.5281/zenodo.18666645 (SAE Foundation Paper 2) References: 10.5281/zenodo.18727327 (SAE Foundation Paper 3) References: 10.5281/zenodo.19329284 (Beyond Fast and Slow) References: 10.5281/zenodo.19305611 (SAE Anti-Turing Test) References: 10.5281/zenodo.19321534 (SAE Psychoanalysis Paper 4) References: 10.5281/zenodo.19366105 (Multi-AI Checks and Balances) Subjects Philosophy of Mind, Epistemology, Artificial Intelligence, Cognitive Science