Leadership in the era of digitisation

On the grey area between man and machine

Source: Management Team (18/02/2020)
Author: Karlien Vanderheyden

In today’s digital world, man and machines are all too often still seen as rivals. However, it is a misunderstanding that applications based on artificial intelligence are set to systematically replace people. On the contrary, we can greatly benefit from making machines our allies. This uncharted territory is what Paul Daugherty and James Wilson (Accenture) call the ‘missing middle’ in their book Human + Machine: Reimagining Work in the Age of AI.

In the past, there was no gap to fill, no ‘missing middle’. Digitisation was mainly used to automate existing processes. Today, sophisticated AI technologies allow for man and machine to join forces. Consequently, the development of the ‘missing middle’ is one of the main challenges organisations face.

People complement machines

People often complement technologies in new jobs that require new skills.

In this framework, one of the new challenges we are confronted with is training of AI systems. A scientist working for a car manufacturer on the latest generation of self-driving cars will need to teach those cars the traffic regulations. Sometimes people break the rules to prevent an accident (they may drive over a continuous white line, for example). These types of ‘human features’ must also be incorporated in the algorithm of a self-driving car.

People are also needed to explain the way complex algorithms work. When a system makes a wrong prediction on the future sales of a certain product, for example, we need to understand the causes and adapt the algorithm.

Last but not least, people are also required to make sure AI systems work properly. When designing new systems, context designers take into account a wide range of factors (user characteristics, cultural differences etc.). General Motors struggled to pinpoint which colour would be most suited to a new robot for the production unit: orange signals danger, yellow is associated with caution... Eventually, the engineers chose a light green shade (safety green). People are also tasked with testing certain values and standards. If an AI system discriminates against people (for example through geographical distance as a parameter in the recruitment of new employees), a person is required to get to the bottom of the issue and rectify the situation.

Machines support people

AI systems can give people exceptional insights based on data that our human brain is unable to process at the same speed. Some organisations use tools to analyse their customers’ feelings during interactions (via Facebook or Twitter, for example). AI systems can also help serve customers much faster, for example through chatbots that instantly answer questions asked by applicants for a vacancy.

AI systems can also support employees physically. In a production unit, for example, a cobot can take on repetitive tasks or lift heavy materials while the operator carries out more delicate tasks. As such, AI systems literally become an extension of the employee’s physical capacities.

What does this mean for leadership?

For a successful collaboration between man and machine, you need to take several aspects into account.

• Adopt the right mindset
  As a leader, it is important that you are open to new working methods. Improvement opportunities may arise internally (e.g. a shorter process to fill vacancies) or externally (e.g. less paperwork to take out insurance). Audi realised that its local technicians were struggling to repair models with complex electronics. They therefore developed robots that are remotely controlled by experts, who work from a centralised location. These robots can help technicians and even communicate with them.
• Experiment
  Setting up experiments helps companies test new products and methods, accelerates the learning process and boosts the improvement process. In Courtrai, Colruyt is experimenting with artificial intelligence to automatically identify fruit and vegetables at the till, which could result in considerable time savings at check-out.
• Build trust
  People are often wary of new technologies, particularly if they see them as a threat to their jobs. As a leader, you can help them feel more at ease when dealing with AI systems. Allow them to test these systems thoroughly and provide sufficient training opportunities. Getting a better insight into what the robot or system “thinks”, e.g. by adding a few dedicated parameters on the dashboard, can also help build trust among employees. People also seem more accepting of new systems when they can adjust the input or control the algorithm themselves.
• Focus sufficiently on data
  The quality of AI systems is highly dependent on the data used. Selecting the right data and analyses is set to become a key task for employees. It is also an essential role in ‘the missing middle’. Last but not least, as a leader you should ensure non-technical staff can also benefit from the information that results from this data.