LATIDIA · Investigación

Comprensión: primero la ingeniería de IA: propósitos, arquitectura de referencia y patrones arquitectónicos para la comprensión humana

arXiv: 2611.13022v1Tipo de anuncio: nuevo Resumen: los sistemas de IA pueden producir resultados confiables sin desarrollar la comprensión humana. Sin embargo, la comprensión humana sigue siendo valiosa para dar forma al trabajo posterior apoyado por la IA, formar

La noticia

arXiv:2611.13022v1 Announce Type: new Abstract: AI systems can produce reliable results without developing human understanding. However, human understanding remains valuable for shaping subsequent AI-supported work, forming preferences, governing collective choices and satisfying curiosity, even when human participation is unnecessary for correcting errors. We propose understanding-first AI engineering, which treats purpose-specific human understanding as a first-class design principle guiding requirements, design and operation. The purposes of understanding are organised into four complementary rationales: reflective agency, effective action, collective governance and epistemic value. A reference architecture connects production, verification, use, and learning and exploration through shared records, supporting understanding through explanation and participation. Five candidate architectural patterns guide choices about parallel learning,