How to Deflate the AI Industry Bubble? Ant Group's Ma Jing Suggests: Improve Asset Utilization and Steadily Seek Real Demand

At the "Taihu World Cultural Forum · Qiantang Dialogue" held in Hangzhou, Zhejiang, from November 12-13, the discussion centered on "Mutual Learning and Innovation in Wealth and Financial Culture." During the forum session titled "Dialogue: Is AI a Bubble or a Revolution?" Ma Jing, Head of AGI Native Applications at Ant Group Co., Ltd., stated that AI is undoubtedly a massive revolution, though it comes with some healthy bubbles.

Ma Jing likened the phenomenon to an aircraft carrier entering the sea, creating ripples, or a rocket soaring into the sky, causing air vibrations—both natural consequences of significant advancements.

She identified two key bubbles in the AI industry: 1. **Technology-Commercialization Gap**: While current AI technology, including chip computing power and energy consumption, can solve many daily problems, the lack of viable business models creates a gap—one of the bubbles. 2. **Product Applicability**: The question of whether AI-driven products address real human needs is critical. Currently, widely adopted AI products, such as conversational assistants like Doubao, robotic vacuums, and autonomous vehicles, remain limited. Many hardware applications, like home companion robots or workplace assistants, have yet to achieve mass adoption.

Ma Jing emphasized that the bubble in AI productization lies in identifying scenarios where intelligent technology can be effectively commercialized. For instance, while AI can perform tasks like cooking or folding clothes, their commercial viability remains uncertain.

Regarding business models, she noted that profitability is fundamental. Technological progress must be guided by commercial viability—finding large-scale, monetizable applications to drive product development. However, the absence of clear financial metrics makes it difficult to assess returns, contributing to another bubble.

To deflate these bubbles, Ma Jing proposed improving asset utilization: - Developing cost-efficient chips. - Optimizing product design. Additionally, she advocated for a long-term, demand-driven approach to AI entrepreneurship, focusing on real market needs to steadily eliminate industry bubbles.