Gartner Group’s 2017 update to their Hype Cycle for Emerging Technologies gives a bird’s eye view on when new technologies might become ready for practical use and increased productivity for individuals and organisations.
At the peak of its hype, any technology is hard to grasp for decision-makers. What is it about? How does it work? Why should I bother? Where does it fit?
For the second time we find Machine Leaning at the “peak of inflated expectations”, with a time-frame of 2 to 5 years to reach mainstream adoption.
Machine Learning fits into operations, i.e., the delivery of products and services. If you manage or deliver products and services, it might be a good idea to take a closer look at how machine learning could optimise the solutions you offer to the market.
Learning machines depend on massive amounts of data to «learn» from, i.e., to be able to identify patterns or situations that can form the basis of decisions or actions. Learning machines depend on stable conditions. If the «rules of the game» change, the machine must be re-trained with new data – data that will take time to collect.
This limitation implies that machine learning does not fit within the area of strategy; an area that is driven by – and aims at creating – change.
Cause-and-effect based simulation is a less hyped – and more promising – technology for investigating future effects of strategic options. Strategy simulations are driven by logic more than data. As a result, simulation is less dependent on massive data, easier to adapt to new insights, and most importantly, simulation opens up for “serious play” where management can experience the likely future effects of different strategic options up front.
In other words: Machine Learning is not a strategic tool. But it can be a good strategic move to introduce Machine Learning at the operational level within your organisation. To prepare yourself and your organisation for the transition, simulation is a good place to start.