In December 2025, Professor Ding Wenxuan, Prof. of AI and Business Analytics from emlyon business school published frontier research in the top-tier management journal “California Management Review”, addressing the critical question: "Why are trillion-dollar investments in generative AI failing to yield returns?" The article, titled "Beyond the Big Data Mindset: An Executive’s Guide to Cultivating AI as Talent" (https://cmr.berkeley.edu/2025/12/beyond-the-big-data-mindset-an-executive-s-guide-to-cultivating-ai-as-talent/), offers a transformative solution.

Since November 2022, enterprises have invested hundreds of billions of dollars in generative AI (GenAI), expecting it to revolutionize their business models. Yet, a staggering number of these initiatives have failed to deliver measurable impacts on profitability. A report from MIT in August 2025 revealed a harsh reality: despite corporate spending on GenAI soaring to nearly $40 billion, 95% of integrated pilot projects have produced no quantifiable return.  

For managers, this is nothing short of a crisis. How can a technology that demonstrates remarkable capabilities in personal applications fail so spectacularly at the enterprise level? While commonly cited culprits—poor data quality, shortage of technical talent, and misaligned strategy—are contributing factors, they point to a deeper, more fundamental cognitive flaw: a persistent "big data mindset" that is fundamentally misaligned with the capabilities of GenAI. We argue that the failure stems from two interrelated core misconceptions:  

• Confusing "Data as Record" with "Processual Know-How"

Enterprises have long mistaken the "data as record"—historical behavioral traces such as transaction logs, user clicks, and performance reports (the "digital exhaust" of past events)—for the valuable "data as know-how"—the often-unwritten processes and methods embedded in production and workflows. This "big data mindset" leads managers to believe that feeding AI more historical data will generate value, while overlooking the need to impart the "process wisdom" that truly drives outcomes.  

• Misguided "Learning" Approaches

We expect AI to "learn" like a strategist—by identifying patterns in historical data to make predictions. Instead, we should teach it to "learn" like a craftsman—by mastering complete workflows. GenAI is essentially a "second mind" capable of understanding, reasoning, and generating content. Its true potential lies not in analyzing the past but in participating in and optimizing the value-creation process itself.

The Solution: From "Deploying Tools" to "Cultivating Talent"

The core issue is not technological but philosophical. Leaders must undergo a fundamental shift in mindset: stop treating AI as a software tool to be implemented and start treating it as talent to be cultivated. We propose a practical, actionable "AI Apprenticeship" framework to guide organizations through this transformation:  

• Step 1: Define the Curriculum

Identify the core business processes and decision logic that AI must master, defining clear "learning objectives."

• Step 2: Cultivate AI as Talent

Enable AI to deeply understand business contexts and operational nuances through interactive training, continuous feedback, and immersive scenarios.

• Step 3: Embed and Grow AI Talent

Integrate cultivated AI capabilities into the organizational structure, establish collaboration mechanisms with human teams, and plan pathways for ongoing development.

GenAI now stands at a critical crossroads—from technological spectacle to genuine value creation. Only by overcoming cognitive misconceptions and completing the paradigm shift from "tool implementation" to "talent cultivation" can enterprises unlock its transformative potential and build enduring, competitive advantages in the age of intelligent revolution.

About Prof. Wenxuan DING

Wenxuan DING

Professor of Artificial Intelligence and Business Analytics
DBA Advisor at emlyon business school

Professor Wenxuan Ding, who holds a PhD in Cognitive Science and Information Technology from Carnegie Mellon University, is currently a Professor of Artificial Intelligence and Business Analytics at emlyon business school, as well as an AI expert affiliated with the French Academy of Sciences and the European Science Foundation. Professor Ding has long been engaged in cutting-edge fundamental AI research and interdisciplinary AI applications, including AI-driven innovation, intelligent healthcare, and risk decision-making. Her scholarly work has been published in top international academic journals and premier conferences in computer science.