Why We Should Ride the Big Data Carriage: Business Perspectives on Data Growth and Machine Learning

Course Name: The Story of Thinking in Machine Learning Instructor: Bi Ran, Baidu Chief Architect Editor: Hoh Xil Source: Machine Learning Training Camp Platform: Baidu Technical Academy, PaddlePaddle, DataFun

Introduction: This session discusses why we should get on the big‑data "carriage" and how data and machine learning can transform business from a product‑owner’s viewpoint, enabling faster growth and stronger market positioning.

Data Growth Curve: Over recent years, both data volume and data richness have been increasing non‑linearly across all industries. As collection methods, IT integration, and mobility improve, more diverse data points (searches, clicks, location, etc.) are gathered, creating value greater than the sum of individual pieces.

Why Ride the Big Data Carriage?

1. Offensive – Promote Business Development: When data volume explodes, business value can also experience exponential growth, linking data growth directly to product and market success.

2. Defensive – Core Competitiveness: Data and data‑driven technology become a company’s core competitive edge, influencing investor decisions and ensuring sustainable returns.

Application Scenarios:

• Personalised matching (search, news, e‑commerce, job matching, ride‑hailing, credit services). • Replacing manual labour with intelligent models (machine translation, facial‑recognition security, automated customer service, autonomous vehicles).

Industry Chain: Data → Model → Business → Demand. Only when these links are fully connected can big‑data truly solve business problems, as illustrated by a shoe‑selling example.

Personalised Matching & Automation:

Examples include restaurant delivery optimisation, personalised education recommendations, and e‑commerce recommendation systems. The internet excels at data collection and execution, making it a fertile ground for big‑data applications.

AI Example – AlphaGo: AlphaGo demonstrates that AI excels in single‑scenario, data‑rich tasks, but cannot replace humans in multi‑dimensional, creative activities. This highlights the "three‑second rule" for assessing which tasks can be automated.

In summary, the session stresses that while technology provides powerful tools, sustainable business advantage comes from integrating data, models, business processes, and real user needs.

For more, follow DataFun and watch the next video in the Baidu Machine Learning series.