Over the past weeks, my focus has shifted from Innovation to something more pressing: keeping the “human in the loop“. This thought resonated throughout several conversations I’ve had this week, where a key concern emerged: How do we retain critical business domain knowledge in an AI-driven world?
The Knowledge Drain Problem
One of the most thought-provoking questions from the TechEx Big AI and Data World Expo was this: what happens when all the domain knowledge leaves an organisation?
Businesses rely on Subject Matter Experts (SMEs). You know them: whenever there’s an issue that can’t be resolved easily, these individuals call on their experience to know which levers to pull. They hold tacit knowledge that’s never been codified.
When they leave, does that knowledge just evaporate? Will future teams have to reverse-engineer the systems just to figure out how the business operates? The reality is that too many companies assume this knowledge is embedded in their processes, only to find it’s walked out the door.
Is AI’s Role Support, or Substitution?
The rise of Gen AI may present an enticing solution for business leaders: let AI fill the gaps. Automatically generate code, design workflows, and handle the complexity. But that itself presents a fundamental risk: without human oversight, who ensures these AI-driven fixes are correct? If we outsource critical thinking to AI, it’s a surefire way for expertise to degrade and diminish over time.
A compelling CIO article highlights this very issue: “With critical thinking in decline, IT must rethink application usability” (link in comments). It argues that over-simplified, “low-friction” applications are encouraging a decline in analytical thinking. Users no longer need to understand the system – they just follow the prompts. But without understanding, how can they challenge, adapt, or suggest improvements for the system?
Balancing Usability with Knowledge Retention
Culturally, we’ve embraced simplicity. Smartphones and minimalist interfaces have conditioned us to expect low cognitive load. But the pendulum may have swung too far. When systems abstract away too much complexity, they deprive users of the opportunity to learn how the business works.
Imagine business applications designed to reveal relationships and processes – systems that help users build a mental model of the business, rather than hiding it. As users interact with the system, they gradually absorb this understanding, reinforcing it through daily tasks. Over time, they become the next generation of Knowledge Stewards – people who understand the business at a deep, system level without needing formal training.
This isn’t about making things harder for the sake of it. It’s about designing tools that encourage exploration and learning, providing clarity rather than obscurity. Low friction, high cognitive connections.
The Need for a Continuous Learning Culture
For this to work, we need more than just better application design. It requires cultural change, driven from the top. Leaders must prioritise a Continuous Learning Culture – one where learning isn’t treated as an afterthought.
Key elements of this culture include:
- AI literacy – Staff need a working understanding of AI: how it operates, where its limitations are, and how to apply critical thinking alongside it.
- Business process awareness – Employees must understand how the business functions, not just their immediate tasks.
- Empowerment to explore – Systems should encourage curiosity, helping users uncover how processes interlink.
One could argue that the time saved by AI automation should be reinvested into deeper learning and understanding. Optimising processes is essential – but genuine learning is a form of business insurance. You don’t think you need it until you do.
Systems Thinking: A Skill for Everyone?
Historically, Systems Thinking – the ability to see how different elements interconnect – was the domain of Engineers, Analysts, and the so-called “IT guy who’s always negative.” In truth, those people weren’t being negative; they were being realistic. They saw the knock-on effects of bad decisions, the vulnerabilities lurking behind oversimplified ideas and solutions.
We need to make Systems Thinking mainstream. It shouldn’t just be the preserve of the technically curious – it should be a mindset instilled across the workforce. When people understand how their role fits into the broader system, they make better decisions, faster. They challenge assumptions, adapt processes, and contribute to continuous business improvements.
Empowering, Not Replacing
The world is changing faster than ever, and organisations must adapt. AI isn’t going away – and nor should it. But its role must be complementary, supporting human expertise rather than replacing it. Usability shouldn’t come at the cost of understanding.
We need to ensure that as AI evolves, so does the workforce – equipped with the critical thinking, business knowledge, and systems mindset necessary to stay in control.
Perhaps I should have attended the talk titled “AI: Terminator or Liberator?“ – because the answer might just depend on how we handle knowledge retention in the age of AI.