The head of ZTE Corporation's (ASX: 000063) consumer business unit, Ni Fei, has offered his perspective on the evolution of AI in smartphones, distinguishing between current models and the next generation. He noted that while many AI phones today offer a growing array of features like smart dialogue and image creation, these are essentially layers of AI functionality added onto the device.
He frankly stated that this approach still leaves users navigating between multiple apps and performing numerous clicks to accomplish a single task. The paradox, he observed, is that the more "capable" the phone becomes, the more "effort" it requires from the user.
Defining the Next Phase
Ni Fei emphasized that product development must address user pain points. He expressed a firm belief that the next stage for AI smartphones will transition from simple "feature stacking" towards becoming a "native intelligent agent." The goal is to shift from users adapting to the phone's logic to having the phone proactively understand human needs, complete tasks, and ultimately provide a more effortless experience.
Core Capabilities of an Intelligent Agent
Outlining this vision, Ni Fei detailed four key capability benchmarks for a true intelligent agent phone. First, it must comprehend natural language without requiring users to adapt their phrasing, and be ready to respond at any time. Second, it must be able to complete assigned tasks from start to finish. Third, it should remember user habits and reminders, eliminating the need for constant repetition and allowing information to accumulate in an organized manner. Finally, and fundamentally, security must form the underlying foundation, with all functionalities built upon comprehensive privacy protection to ensure user peace of mind.
Upcoming Product Showcase
Previously, Ni Fei confirmed via social media that the company's Nubia brand will launch an AI agent smartphone. This device, described as a mass-produced flagship model, is scheduled for its first public demonstration at the 2026 World Artificial Intelligence Conference.
Comments