There is much talk of Artificial Intelligence (AI) and its role in industry to automate certain job functions, in turn driving organisational efficiencies and ROI. And it’s not just about replacing those in low wage, low skill jobs, increasingly we are hearing it’s actually the well remunerated, judgement-based roles that are most financially attractive targets for AI. After all, setting up AI isn’t cheap. Cheap labour won’t cover the costs—the reduced expense of fewer middle managers and sales people just might.

When it comes to learning and development, AI is already being used to gather data from learners or from people doing the job someone is training for. Using algorithms, software can accurately recommend possible learning topics and serve up appropriate content before employees know they need it.

For those working in the field of sales, in theory at least, this could mean teaching sales teams to be more persuasive, behaviourally subtle, and capable of moving a conversation towards a positive customer commitment. But can it deliver?

Simply telling people to do things isn’t enough when we want to help them to adopt high performance behaviours.  They need practice, feedback and opportunities to reflect on their experience in order to create an informed action plan for the next call and the next customer. That’s not serving them up yet more digital content—that’s a concerted effort to facilitate behaviour change. It needs people. There is no technological quick fix. More content—however well targeted—is a blind alley not a road to a new future.

This rush to AI in learning is not without other challenges.

To work, AI requires a wealth of data. There is a real danger of rushing to crunch whatever numbers we can get hold of, and as a result, we run the risk of magnifying that tendency—valuing the things that can be counted instead of really important things that are more difficult to quantify. For those undertaking sales training, the data generated may be quite limited. The system may know which course was taken and when.  Scores from online courses or details of tools and resources which have been downloaded might be available. But it’s pretty small in terms of the data required to power an AI driven learning system.

There’s a real risk that organisations will try to implement big data actions on small data sets.

By using AI to determine learning needs, it can potentially reduce the range of options available and pushes everyone to a kind of ‘learning magnolia’—a safe, bland alternative which limits our experience rather than expands our horizons.

It is, in a real sense, like driving a car by only ever looking in the rear view mirror. ‘People like you also completed this online module’ is barely relevant in an environment seeking high performance. It doesn’t take into account, people’s strengths, weaknesses and interests or passions.

This use of past data ignores the most important part of learning: that learning is transformative. Good learning experiences change us. The application of AI to predict future needs based on past activity risks not delivering on this. Instead, effective learning has to be delivered through a rigorously researched, timeless and validated methodology based on significant data sets obtained via observation of successful people—not counting the clicks of the few to determine a strategy for the masses.

By Robin Hoyle, head of learning innovation, Huthwaite International