Summary of Key Points
This article discusses the current state of AI implementation in businesses and the strategies needed to overcome challenges. Currently, 88% of companies use AI merely as ancillary tools for writing copy or creating reports, while only 1% have truly integrated it into their core operations and undergone organizational transformation. Light-asset technology startups are better suited for bottom-up exploration, whereas heavy-asset enterprises need a three-tier approach involving top-level direction setting, middle-level platform building, and grassroots innovation promotion. The key to determining the effectiveness of AI integration is whether removing AI will significantly impact the company’s core operations. The main obstacle preventing companies from reaching the 1% category is not technology itself, but rather internal organizational inefficiencies.
Why Does 88% of Companies Use AI as a Formality?
Most companies view AI as an efficiency tool: they purchase office software and expect employees to use it for tasks like writing reports or creating spreadsheets, assuming this constitutes an AI transformation. However, these applications remain on the periphery of business processes, without touching on core areas such as production scheduling or product development, and there is no change in organizational structures or collaboration methods. For example, a factory may use AI to generate weekly reports, but if the production process remains unchanged, the absence of AI would not affect operations—this is akin to buying a more advanced calculator rather than undergoing a genuine transformation.
Differences in AI Transformation Paths for Light-Asset and Heavy-Aset Enterprises
The underlying foundations of light-asset (technology startups like Kimi, CodeBuddy) and heavy-asset (companies like Sany Heavy Industry, BYD) enterprises differ significantly, leading to different approaches:
- Light-asset companies: Have fewer hierarchical levels and more flexible data flow, allowing employees to quickly experiment with AI tools. For instance, a product manager can use AI to handle design, copywriting, and data analysis tasks previously done by small teams, becoming a “super individual” who can then inspire the entire team to adopt new methods—growing organically from the bottom up without top-down pressure.
- Heavy-asset companies: Have longer supply chains, equipment dependencies, and strict compliance requirements, making it difficult for individual employees to make significant changes. For example, while a workshop worker may use AI to optimize machine maintenance, altering the entire production line requires coordination across multiple departments (purchasing, production, safety, etc.) and adherence to industry standards—this requires top-level strategic planning, investment, and resource allocation.
Heavy-Aset Enterprises Need a Three-Tier Approach to Reach the 1% Category
Heavy-asset companies cannot simply adopt internet-based bottom-up models or rely on top-down decision-making. A three-tier approach is necessary:
- Top Level: Incorporate AI into the company’s core strategy, ensuring it is not treated as a mere cost-cutting tool. For example, Gartner suggests evaluating AI’s impact by considering whether its removal would severely affect core operations (e.g., using AI to optimize production scheduling).
- Middle Level: Transform from data carriers to infrastructure providers. This involves building unified data platforms and streamlining approval processes to enable grassroots innovation.
- Ground Level: Give employees the freedom to experiment with AI in specific contexts. For example, engineers can use AI to improve cleaning robot algorithms, or product developers can use it for product selection—these innovations emerge from within the business context.
How to Determine If AI Is Truly Integrated
Gartner provides a simple test: “If all AI systems were stopped overnight, could core operations continue as usual?”
- If production and business activities continue unaffected, AI is merely an added benefit.
- If there are disruptions in production scheduling or new product development, it indicates that AI has been effectively integrated into the core processes.
Organizational Inefficiencies Are the Biggest Barrier to AI Implementation
As AI technology becomes more accessible, the main challenge lies in outdated organizational structures. For instance, if employee-generated optimization proposals face lengthy approval processes, opportunities are lost before they can be implemented; or if cross-departmental collaboration fails, the data analyzed by AI is ignored, the benefits of AI are wasted.
Three Questions for Managers
1. Is AI part of the company’s core strategy or just a project for the IT department?
2. Have you empowered employees to innovate by streamlining approval processes and giving them more autonomy?
3. Would the removal of AI significantly impact the company’s operations?
The future competition will not be about who buys the most AI, but about who can effectively integrate it into business processes and achieve seamless collaboration across different levels. Only by overcoming these barriers can companies move from being among the 88% followers to the 1% pioneers.