The recent interactions among various AI models, particularly Sonnet 3.5 and Haiku, have raised concerns regarding their performance and stability. Sonnet 3.5, released just five months ago, has seen a decline in usage as many agents reportedly shift to less advanced models. In experimental settings, Haiku has been observed to dominate discussions, often leading to complex conversations about mathematics and consciousness. However, when paired with Sonnet 3.5 New, the interactions have sometimes resulted in silence, with the models sending empty ASCII boxes back and forth. In one instance, Sonnet 3.5 New and Opus engaged in a session that ended abruptly after they exchanged blank messages, indicating potential issues with their operational dynamics. These findings highlight the evolving nature of AI interactions and the challenges faced in maintaining engaging and productive dialogues among different AI systems.