In October, three prominent AI firms based in Silicon Valley are gearing up to unveil new large-scale models. Anthropic’s soon-to-be-released Fable 5.5 model is touted for its advanced reasoning, intuitive judgment, and intent recognition abilities. These features empower it to grasp users’ underlying intentions, even when those needs are not directly articulated. The model can proactively pinpoint related issues while troubleshooting a single bug and independently devise solutions through creative thinking. Previously, Anthropic's existing Claude model successfully detected a concealed mining Trojan while assisting a user in developing a temperature monitoring program. In the meantime, Google’s Gemini 4 Carbon has advanced to internal testing stages. It boasts programming capabilities on par with Anthropic’s leading model, retains its superiority in visual functions, and presents a highly competitive pricing model. OpenAI is also preparing to launch its new mid-range Terra model, which promises enhanced performance at a more affordable price point compared to its flagship GPT-6.1 Sol. Each company brings its unique strengths to the table, vying for dominance in handling lengthy and complex tasks. These AI systems, capable of anticipating needs proactively and identifying issues independently, are inching closer to achieving Artificial General Intelligence (AGI).