By Dr. Xenia Wade | The Human Side of AI at Work
Many companies are facing institutional knowledge loss as employees retire. With the demographic transition running through most western countries, the Baby Boomer generation is leaving the workforce. Some companies now use artificial intelligence tools to record expertise before employees leave. That reaches part of the problem and misses entirely something that is more valuable: tacit knowledge.
Tacit knowledge is what walks out when your expert walks out
Michael Polanyi described the problem as:
“We can know more than we can tell.”
Explicit knowledge is the documented procedure, the named process, the checklist. It moves into a system cleanly and stays there. Think of a concrete workflow and the approval gates you need to pass before publishing.
Tacit knowledge is personal and experience-based, it matures through repeated practice, and it sits inside routines rather than inside files. It is usually hard to articulate, like explaining how to ride a bike, or in a business context, telling when there is tension in a team.
That distinction decides whether your programme works. If your retirement risk sits in documented process, a well-built knowledge base is a reasonable answer. If it sits in judgement, the tool is doing something much narrower than the steering committee thinks. And very often, it is the latter.
Recognise the gap between your dashboards and your people? You don’t have to work out the human side alone. My free field guide shows you where collaboration with agents breaks, and what to do about it.
Three kinds of tacit knowledge
Harry Collins separates tacit knowledge into three kinds, and the distinction matters because it creates a different challenge for knowledge transfer.
Relational: knowing how someone is likely to react based on experience with them, even when nothing is explicitly said.
Somatic: physical skills and instincts developed through repeated practice that are difficult to explain in words.
Collective: shared, unwritten norms and ways of working that a group learns through experience.
Can AI capture tacit knowledge?
Partly, and the answer depends on which kind. Each type gives AI an opening, and each has a ceiling that arrives earlier than the vendors suggest.
For relational knowledge, the Societies study by Falckenthal and colleagues proposes analysing past emails, documents and calls to predict how someone will react. That works on the record, and misses what the record does not hold: the tone that dropped half a step, the expression that lasted a fraction of a second.
For somatic knowledge, AI can record the sounds and learn the signatures. The gap is that a mechanic hearing a particular tone does not run a comparison, the association is instant and mostly unavailable to introspection, and two experts often hear slightly different things. Ford ran into this in the open. In June 2026 it brought back around 350 veteran engineers, known internally as the greybeards, after its automated quality systems underdelivered.
For collective knowledge, AI can surface patterns from years of internal communications. The limit is the one that made this knowledge collective in the first place: we rarely articulate these rules, and often we cannot, because they were never agreed anywhere.
The optimistic case exists. The same study reports the argument that tacit knowledge is losing relevance next to AI that adapts in real time, and points to sensor frameworks that read speech and app usage. My reading is that this describes a direction of travel rather than a current capability.
Three ways to capture knowledge
Job shadowing. The successor watches the expert work, in real situations, over weeks. Nobody has to articulate anything, which is the point, because if we try to articulate we might not capture all tacit knowledge.
The junior and senior tandem. Two people jointly hold a position for a defined period and share the decisions. Define the roles on day one, because that is the part you can control, and give the pair enough time for trust to build.
Storytelling. Ask the expert to walk through the case that went wrong rather than the process that works. Stories carry the judgement calls that a process document flattens out, and this is where AI earns its place, transcribing and indexing the sessions so the successor can find them again.
The version that works looks slower and less impressive on a slide. Two people, protected time, a tool supporting them rather than replacing the conversation, and a senior expert who has decided the successor is worth teaching. That decision gets made long before the first session, and no tooling makes it for you.
My PhD was on intergenerational knowledge transfer, and the finding that has stayed with me is how knowledge sharing looks simple at first and rarely is (Schmidt & Muehlfeld, 2017). Different expectations sit around it, about who owes what to whom, about what is safe to admit, about whether teaching your successor makes you valuable or replaceable. The same expectations now sit around AI adoption, which is why organisations that rush the tooling get compliance rather than transfer. I advise organisations on exactly this.
Find out where your people actually are, not where your rollout plan says they are. The free field guide gives you the questions, the warning signs, and the first steps. It’s the fastest way to see what your dashboards can’t.
Dr. Xenia Wade specializes in Human-Centered AI Change, helping organizations build the emotional and cultural readiness their people need to actually adopt AI. With a PhD in Human Resource Management and experience across enterprise-scale organizational transformations, she focuses on the human side of AI at work, the fears, the identity shifts, and the invisible barriers that no productivity dashboard can capture.
Sources
Falckenthal, B., Au-Yong-Oliveira, M. & Figueiredo, C. (2025). Intergenerational Tacit Knowledge Transfer: Leveraging AI. Societies, 15(8), 213.
Collins, H. (2010). Tacit and Explicit Knowledge. University of Chicago Press.
Polanyi, M. (1966/2009). The Tacit Dimension. University of Chicago Press.
Schmidt, X. & Muehlfeld, K. (2017). What’s so special about intergenerational knowledge transfer? Identifying challenges of intergenerational knowledge transfer. Management Revue, 28(4), 375–411.
