Unlocking Unwritten Rules: Teaching GenAI To Think Like An Expert

Forbes

27-03-2025

Rajesh Rajagopalan, Co-founder, CEO & CTO of PeerIslands.
Consider how someone develops expertise in the workplace. Beyond formal training, new employees learn unwritten rules, insider tips and valuable knowledge from experienced colleagues.
In software engineering, for example, developers learn debugging techniques, preferred design patterns or the quirks of legacy systems. In other fields, support specialists identify high-maintenance clients (for example, mechanics understand machine idiosyncrasies, and legal analysts learn which regulations require extra scrutiny). This accumulated wisdom is often called "institutional knowledge" and sets top professionals apart.
This deep understanding, intuition and wisdom distinguishes true masters from those merely following instructions. It differentiates average performers from "10x" engineers, marketers, service agents or nurses.
However, developing and preserving deep expertise is a growing challenge, with Baby Boomers retiring and remote work disrupting traditional mentorship models.
GenAI provides a solution by capturing, preserving and scaling this knowledge. Instead of replacing human expertise, GenAI serves as a powerful tool for learning from patterns, experiences and decisions—replicating the "gut instinct" and insights of seasoned professionals.
Here, I'll explore how GenAI can help organizations retain and transfer critical knowledge, ensuring expertise remains accessible in a rapidly changing workplace.
To appreciate why institutional knowledge has been so difficult to document or automate, let's look at how it manifests in different industries.
In manufacturing, experienced operators can detect potential equipment failures through subtle changes in vibration or sound that novices might overlook. Similarly, seasoned nurses integrate subtle patient cues in healthcare to anticipate deterioration risks long before vital signs change.
This expertise is largely tacit—deeply intuitive, context-dependent and difficult to articulate, document or transfer. It relies on complex pattern recognition and decision-making that traditional knowledge management systems struggle to capture. Conventional, rule-based approaches often fall short of preserving and passing on these nuanced insights.
Organizations have long used documentation, knowledge bases and formal training to capture and transfer expertise. However, these methods have critical shortcomings:
• Documentation quickly becomes outdated. Written materials struggle to keep up with evolving processes and often fail to capture the deeper reasoning behind expert decisions.
• Traditional knowledge bases lack accessibility. Information is frequently scattered across different systems, making it challenging to locate, retrieve and apply in daily workflows.
• Training is limited and inefficient. Formal training sessions occur sporadically, take time away from work and primarily focus on explicit knowledge rather than the practical experience that drives sound decision-making.
These methods also heavily burden experts, who must step away from their core responsibilities to document and explain their thought processes.
Generative AI is transforming how expertise is captured and shared. Instead of relying on manual documentation, GenAI can learn directly from data and interactions, making knowledge management more efficient.
For example, by analyzing an engineer's past code reviews, GenAI can learn their coding style, design choices and quality standards. By processing a nurse's clinical notes alongside patient outcomes, GenAI can uncover the subtle cues they use to anticipate complications.
This approach offers several key benefits:
• Learning From Unstructured Data: It can extract expert knowledge from notes, emails, cod, and conversations.
• Pattern Recognition: AI can analyze thousands of past decisions and uncover insights that even experts might overlook.
• Context Awareness: Unlike traditional systems, GenAI integrates multiple data sources to provide a deeper understanding of situations.
• Continuous Learning: AI improves over time, adapting to new information and recognizing when human expertise is needed.
• On-Demand Knowledge: GenAI delivers insights within workflows, ensuring knowledge is available when and where it's needed.
In essence, GenAI acts as a smart and tireless apprentice—learning from experts, refining its understanding and making valuable knowledge easily accessible.
As GenAI matures, we can expect even more transformative impacts on knowledge work:
• Personalized Learning: GenAI could create customized learning paths for each employee, recommending experiences and mentors to accelerate expertise development.
• Collaborative Intelligence: GenAI could facilitate collaboration by matching employees with complementary expertise and translating knowledge across domains.
• Organizational Memory: GenAI could preserve critical knowledge even as employees move on, becoming a repository of an organization's collective wisdom.
• Innovation Catalyst: By uncovering novel patterns across diverse expertise domains, GenAI could suggest innovative solutions and research directions.
Ultimately, GenAI could enable organizations to treat expertise as a renewable strategic asset rather than a scarce, depleting resource.
In my experience, realizing GenAI's potential for knowledge management requires addressing several challenges:
• Explainability: GenAI systems must be able to explain their recommendations for experts to understand and validate them.
• Bias And Fairness: GenAI models must be audited for biases that could lead to discriminatory or harmful recommendations.
• Privacy And Security: Robust data governance is needed to protect sensitive expert knowledge and employee privacy.
• Human-AI Collaboration: Workflows and interfaces must be designed to facilitate smooth cooperation between experts and GenAI.
Building trust is essential. Experts should view GenAI as a tool to expand their expertise, not a replacement. Involving experts in GenAI systems' design, training and governance is critical for encouraging adoption.
I have been fascinated by J.A.R.V.I.S. from Iron Man and the chip's computer from the Star Trek series—intelligent systems that seamlessly assist, enhance and extend human capabilities. Integrating GenAI into knowledge management brings us closer to this vision, offering a powerful opportunity to enhance, rather than replace, human expertise.
By learning from experienced professionals, GenAI can help organizations capture, distribute and apply critical knowledge, enabling experts to focus on next-level challenges.
While challenges remain, organizations that navigate this shift successfully can create an environment where knowledge flows seamlessly and the partnership between human and artificial intelligence fuels continuous innovation. This is not just about preserving expertise—it's about elevating it to new heights.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?