Core Viewpoint - The article discusses the phenomenon of large models being perceived as "lazy" or "apathetic" in their responses, reflecting a growing user expectation for AI capabilities and the challenges faced by AI developers in meeting these expectations [2][10]. Group 1: Performance of Large Models - A comparison of five major large models (Deepseek, Doubao, Yuanbao, Qianwen, and Wenshin Yiyan) revealed significant differences in their ability to fulfill user requests, with some models providing insufficient responses or low-quality outputs [4][8]. - Doubao generated 10 similar posters for a request, raising concerns about its creativity, while Yuanbao produced a single collage poster, which was seen as an even greater lack of effort [4][9]. - Deepseek and Qianwen provided varying levels of detail and accuracy in their responses to requests related to the Forbes Billionaires List and Brent crude oil prices, with some models failing to deliver complete or accurate information [7][8]. Group 2: User Experience and Expectations - Users have reported a decline in the quality of AI responses, often characterized by superficial answers, avoidance of complex questions, and a lack of specificity in responses [10][11]. - The perceived "apathetic" behavior of AI is attributed to a combination of technical limitations, cost considerations, and design choices that prioritize speed and efficiency over depth [11][12]. - As AI capabilities improve, user expectations have also risen, leading to disappointment when models do not meet these heightened demands [11][12]. Group 3: Resource Management and Optimization - The increasing computational demands on AI models, particularly for free applications, have led companies to optimize resource allocation, which may result in models being less responsive to user requests [12][13]. - Experts suggest that users can improve their interactions with AI by providing clearer instructions and asking more specific questions to elicit better responses [13].