https://cuicaihao.github.io/tags/generative-ai/Recent content in Generative AI on C.CUI's LogHugoen-AUSat, 22 Aug 2026 07:00:00 +1000https://cuicaihao.github.io/posts/2026-08-22-why-an-ai-animated-series-can-reach-eleven-seasons/Sat, 22 Aug 2026 07:00:00 +1000https://cuicaihao.github.io/posts/2026-08-22-why-an-ai-animated-series-can-reach-eleven-seasons/Generating a single shot tests model capabilities, but continuously updating a series for eleven seasons around the same characters and worldview requires a complete production system. Using the Chinese AI anime and short-drama industry as context, this article analyzes how task breakdown, production assets, model orchestration, agentic collaboration, human decision-making, and recommendation feedback sustain continuous delivery. As generation costs decline, competitive advantages will likely shift from models to systems.https://cuicaihao.github.io/posts/2026-08-19-how-generative-ai-is-reshaping-recommendation-systems/Wed, 19 Aug 2026 07:00:00 +1000https://cuicaihao.github.io/posts/2026-08-19-how-generative-ai-is-reshaping-recommendation-systems/This article traces the evolution of recommendation systems from collaborative filtering and deep learning to the transformative impact of generative AI. It delves into how generative AI alters user interaction through natural language, integrates cross-application context, and participates in content organization and generation, moving beyond traditional retrieval and ranking. The post further distinguishes between current product implementations and research prototypes, discussing their implications for user choice, content production, and information verification.