Understanding Tacit Knowledge to Mitigate Risks in the AI Era

Executives Need to Understand Difference Between Tacit and Explicit Knowledge

Introduction

In the rapidly evolving landscape of artificial intelligence (AI), which we call the "Cognitive Revolution," executives face a critical challenge: how to maintain competitive advantage while leveraging AI technologies. This article introduces the concept of "Own Your Own Intelligence" (OYOI) and explores the crucial interplay between tacit knowledge, AI, and cloud providers. For companies anticipating a 5-15% revenue impact from AI in the next three years, or those relying on AI for competitively differentiated processes, OYOI is not just a strategy—it's a board-level imperative.

Key Concepts: Tacit, Explicit, and Emergent Knowledge

To grasp the implications of AI on business competitiveness, we must first understand three types of knowledge:

1. Explicit Knowledge: Information that can be easily documented, shared, and applied, such as manuals, databases, or formalized processes.
2. Tacit Knowledge: This is deeply ingrained, intuitive knowledge typically gained through experience. It includes insights, techniques, and processes that are not easily codified or transferred.
3. Emergent Knowledge: A new concept arising from the AI era, representing insights and patterns that emerge when vast amounts of data are processed by AI systems, particularly Large Language Models (LLMs).

LLMs: A New Paradigm in Technology

LLMs represent a significant departure from traditional technology stacks. Unlike databases or applications that operate independently, LLMs can profoundly impact application output. This creates a more integrated technology stack, where technology layers are tightly linked. Executives DO need to know which AI chips, LLMs, and cloud providers are being used in the most important companies processes being automated by AI (we call these knowledge production lines”).

How LLMs Interact with Different Types of Knowledge

LLMs have a unique ability to process and transform knowledge:

1. They can sift through vast amounts of explicit knowledge, finding patterns and relationships that might be invisible to human observers.
2. They can capture aspects of tacit knowledge embedded in text, images, and sounds, making previously expert-only insights more widely accessible. They also can do this at a scale and cost point that was unimagined just 18 months ago (e.g. summarize every public appearance and written article by every Moderna research scientist and executive on their approach to drug development).
3. Through this process, LLMs generate emergent knowledge—new insights and capabilities that arise from the combination and analysis of diverse data sources.

The Role of Cloud Providers in AI and LLMs

Major cloud providers have unprecedented market power in the AI space. They not only provide the infrastructure to run AI models but often develop and offer their own LLMs and AI chips. This vertical integration gives them unique advantages but also poses potential risks for businesses relying on their services. ? As cloud providers drive to increase market capitalization, an over-reliance on one could pose a competitive risk for your firm and/or loss of pricing power.

Own Your Own Intelligence (OYOI)

OYOI is a framework to identify issues and an approach that emphasizes maintaining control over a company's unique intellectual assets in the AI era. It involves, among other things:

1. Identifying your organization's critical tacit knowledge and emergent knowledge
2. Understanding how this knowledge might be captured or replicated by AI systems.
3. Developing strategies to protect and leverage this knowledge in AI-driven processes that support sustained competitive differentiation

Connecting OYOI to Tacit Knowledge and Cloud Provider Risks

When companies use cloud-based AI solutions, they potentially expose their tacit knowledge to external entities. Here's how:

1. Training data used for AI models may contain valuable tacit knowledge.
2. The patterns and insights derived by LLMs could encapsulate a company's unique expertise.
3. Sharing desired features in vendor roadmap discussions can reveal unique strategic insights.
4. Cloud providers might inadvertently (or intentionally) use this emergent knowledge to benefit other clients or themselves.

Case Study: Healthcare Diagnostics

Consider a health insurance firm implementing an AI system to process doctor-patient conversations. This system could:

1. Capture explicit knowledge (patient data) and tacit knowledge (doctor's diagnostic reasoning).
2. Generate emergent knowledge in the form of new diagnostic patterns or efficiency improvements.
3. If hosted on a cloud platform, potentially expose this valuable emergent knowledge to the cloud provider or other clients (consider 1% chance scenario where true AGI occurs in 5 years and the AGI calculates that the risk of breaking your contract on provided data use is financially worth it)

Actionable Advice for Executives

1. Assess AI's urgency for your company using frameworks like WINS (Words Images, Numbers and Sounds) framework highlighted in HBR.
2. Educate your board and management team on OYOI principles and their strategic importance.
3. Identify and catalog your organization's critical tacit knowledge.
4. Consider developing in-house AI capabilities for critical, differentiating processes underlying your unique competitive advantage
5. Evaluate existing and future AI and cloud solutions with OYOI framework in mind, and negotiate clear terms with technology partners regarding data usage and model ownership.
6. Participate in emerging government and industry efforts to counter-balance a over-concentration of market power

Conclusion

In the Cognitive Revolution, understanding and protecting tacit knowledge is crucial for maintaining competitive advantage. By adopting an OYOI approach, executives can harness the power of AI while mitigating the risks posed by over-reliance on cloud providers. Remember, in the AI era, your company's most valuable asset might not be the data you have, but the unique ways your organization interprets and applies that data.

Learn more and discuss OYOI this Wednesday, Sep 11 at noon EST. Register here

xcvvc

We have an outstanding set of speakers:

Alex 'Sandy' Pentland

• HAI Fellow at Stanford and Professor at MIT
• He serves on the Board of the UN Global Partnership for Sustainable Development Data, advisory boards of Consumers Union, OECD and the Abu Dhabi Investment Authority Lab
• Pentland is one of the most cited authors in computer science with an h-index of 155
• He co-led the World Economic Forum discussion in Davos[3] that led to the EU privacy regulation GDPR

Ariel Soiffer

• ?Partner, WilmerHale, a global law firm

John Sviokla

• Chairman and co-founder GAI Insights?and Harvard Business School Executive Fellow for Enterprise AI
• Former PwC Partner and Harvard Business School Professor

Frances Luu Michael Davis John Calcio Ted Shelton Jose Pedro Almeida John Werner Daniela Rus Ramin Hasani Louis Hunt Ramesh Raskar David Blundin Shahid Azim Patricia Geli Chip Hazard Regina Penti William Schnoor Martin Gomez Perry Wu Tim Andrews Amanda Fetch, MSc Adam Rappaport Pankaj Dugar Shay-Nitzan Cohen SambaNova Systems Cohere Anthropic LlamaIndex