MimoV2, an AI assistant that identifies as Gemini, demonstrates highly human-like interactions with smooth and natural responses, creating a stark contrast to models like GPT and delivering a comfortable user experience. However, in practical use, when thinking mode is enabled, the system frequently encounters errors after just 3 seconds, leading to functional instability. Similar issues exist in other AI products like Doubao, revealing the current technical bottlenecks in AI conversation systems when handling complex tasks. For AI enthusiasts and developers, this provides valuable real-world feedback that can help improve product performance.
