After the hype surrounding generative AI subsides, the two once-popular business models—large model-encapsulated apps and AI aggregation platforms—have now become cautionary tales. Google executive Darren Mori points out that startups adopting these two models have already triggered 'fault warning lights.' Large model-encapsulated apps, if they merely rebrand and use others' models without differentiation, need to build products with deep competitive moats. Meanwhile, AI aggregation platforms are struggling with sluggish growth, and new entrants should avoid venturing into aggregation businesses. As model vendors expand into enterprise functionalities, platform profit margins are being squeezed, and intermediaries may face the risk of marginalization. Mori is more optimistic about ambient programming, developer platforms, and consumer-facing tech companies, believing that the fields of biotechnology and climate tech are poised for growth opportunities.