Lori Niles-Hofmann, EdTech + AI Transformation Strategist and Analyst and author of The Eight Levers of EdTech Transformation argues for a shift in organisational learning.
There is a tension running through corporate learning that we rarely acknowledge out loud. We celebrate curiosity. We encourage continuous development. We design ever more pathways, programmes, and platforms to help people learn. And yet, the people we are asking to learn are under immense pressure. Their roles are changing. Their workloads are expanding. Their attention is fragmented. Learning, in practice, behaves less like a gift and more like a tax on time, energy, and cognitive bandwidth.
This is not a moral failing on the part of employees. It is a systems problem. We have built learning functions that assume infinite capacity to absorb change, while the reality of work keeps getting heavier. At some point, the maths simply stops, well, mathing. No amount of inspirational messaging about growth mindsets compensates for the fact that people still have jobs to do.
To keep people curious is noble. To keep people employed is divine.”
This is why curiosity alone is not enough. To keep people curious is noble. To keep people employed is divine. That shift in emphasis changes how we think about learning altogether.
For years, the dominant response to disruption has been ‘more upskilling’. More content. More courses. More learning journeys. More time spent away from work to prepare for work. Upskilling became a virtue in and of itself, largely disconnected from questions of timing, relevance, or return. In practice, this often resulted in well-intentioned overload. We asked people to learn everything, just in case they might need it one day.
Rightskilling starts from a different premise. It assumes that learning capacity is finite and therefore precious. It forces us to be intentional about what we ask people to learn, when we ask them to learn it, and why. Not every skill gap deserves a learning intervention. Not every capability needs to be built internally. And not every future scenario warrants preparation today.
The supply chain problem
In my book, The Eight Levers of EdTech Transformation, I describe skills as a supply chain problem because it is one of the few metaphors that captures the operational reality of what organisations are already doing, albeit poorly and without visibility. Skills enter the organisation through hiring, learning, and experience. They decay. They become obsolete. They sit unused. They spike in demand with little warning. Most organisations have no real-time view of what they have, what they are short on, or how long it would take to close the gap.
To be clear, calling skills management a supply chain does not reduce people to inventory. It acknowledges something far more sobering: running a business without understanding your capability pipeline is reckless. Organisations already make sourcing decisions about skills every day. They just do it implicitly, inconsistently, and without data. The cost of getting this wrong is not theoretical. It shows up as burnout, redundancy, stalled strategy, and people being trained for roles that quietly disappear.
To see what this looks like when done well, it helps to look at a modern, digital-native supply chain. Consider Amazon’s handling of something as unremarkable as greeting cards. Amazon does not attempt to design, print, and warehouse every possible card variation just in case someone might want it. Instead, it makes constant, data-informed decisions about what to stock, what to source from third parties, what to produce on demand, and what to replace entirely with a digital alternative. Inventory shifts with seasonality, geography, and demand signals that update continuously. When demand drops, products disappear without ceremony.
Some skills should no longer be bought or built. They should not even be borrowed. They should be automated.”
This is the level of precision most organisations lack when it comes to skills. We talk about “needing more AI capability” or “building digital skills” as if those statements are meaningful on their own. They are not. What matters is knowing whether you need 16 people with a particular capability or 17. That difference determines everything. Including whether learning is the right lever at all, urgency and cost. It determines whether you build, buy, borrow, or do something else entirely. There is now a fourth option that must be taken seriously: Bot (thanks to Annee Bayeux for introducing this model).
Some skills should no longer be bought or built. They should not even be borrowed. They should be automated. This is not a statement about the value of people. It is a statement about the value of their time and the importance of efficiency. Data extraction, summarisation, scheduling, first-pass analysis, reporting, and signalling tasks increasingly fall into this category. Training humans to perform these bot tasks deeply is not an act of respect. It is an avoidance of harder decisions to rethink workflows entirely. This type of thinking also frees up humans to do the type of work that is high-value and meaningful.
Creating the eco-system
None of this is possible, however, without a proper backbone. In The Eight Levers of EdTech Transformation, I argue that skills only become useful when they sit inside an ecosystem, not a standalone platform or a static taxonomy. An ecosystem connects skills data to the systems where work actually happens: HR, talent, learning, project management, performance, and increasingly, AI agents operating across those environments. Without this connective tissue, skills remain descriptive. Interesting, perhaps, but inert.
Now, there are endless debates about whether skills are the ‘right’ unit of measurement. These tend to be a distraction in my opinion. The unit matters far less than the discipline of measurement and the courage to act on it. You could model tasks. You could model competencies. You could model outcomes, pints, kilograms or yards! What matters is that you choose something, track it intentionally, and use it to make decisions that change investment patterns and behaviour.
Curiosity and growth still matter. But they must be anchored in relevance, timing, and constraint.”
Rightskilling is not a softer version of upskilling, it is an extension of it with more rigour. It asks HR and L&D leaders to stop broad-stroke sheep-dip programmes and pathways, and instead, optimise for activity, outcomes and usefulness. To accept that saying no to learning is sometimes the most responsible decision you can make. Afterall, our first leader to serve in L&D is the learner, after the business. We need to advocate for them.
To be clear, curiosity and growth still matter. But they must be anchored in relevance, timing, and constraint. In a world where work is accelerating and capacity is shrinking, restraint is not a lack of ambition. It is a signal that we better understand the system we are responsible for.
