On April 12, during the High-Level Forum on the Development of Intelligent Electric Vehicles, Zhou Guang, the CEO of AutoX, made a prediction. He stated that by 2025, the user retention rate for NOA (Navigate on Autopilot) in urban areas is projected to fall between 20% and 30%. Meanwhile, the market penetration rate of NOA in these regions is expected to hover around 15%. Zhou Guang highlighted that the fundamental reason behind this disparity lies in the notably limited number of parameters in the currently mass-produced small-model intelligent driving systems. The majority of these systems boast parameters below 1 billion, or even as low as 0.1 billion, with operational computing power typically ranging from 100 to 200 TOPS (Tera Operations Per Second). To counteract the 'seesaw effect' observed in small models, it is imperative to transition from a small-model framework to a large-model paradigm. This shift will enable a systematic enhancement of capabilities through scaling up.