Why computers should select business leaders – Otago study
FYI, this story is more than a year old
In the near future, machines are going to have a say in who we select as leaders in organisations – an Otago-led study calls for caution in this brave new world.
Lead author Brian Spisak, of the Department of Management, says machine learning (the use of artificial intelligence to develop computer programmes that access data and use it to learn for themselves) is powerful and will play an increasingly larger role in all of our lives.
“Basically, where there’s data, there’s going to be some sort of machine learning algorithm exploring patterns.”
“It will shape how we select leaders (and followers), build teams, make decisions at a strategic level, adjust patterns of behaviour in real-time, dictate marketing plans, and adjust how we distribute budgets and make investments,” he says.
Businesses and organisations already use machine learning to aid in the hiring process, but they also hold masses of other untapped data, particularly related to leaders.
“We realised it’s only a matter of time until the mountain of leadership personality data is used to predict leader effectiveness. It’s increasingly important to better understand what machine learning can add as well as what it can’t. How will be people feel about a machine judging whether they have what it takes to become a CEO, for example? How much should we rely on the machine?” Spisak asks.
The study, just published in The Leadership Quarterly, used self-report personality data and performance evaluations of 973 managers from a range of areas. The researchers used machine learning to investigate the theory of different personality traits being linked to successful leadership.
They found personality predicted performance and, importantly, the context of the situation, significantly improved the ability to calculate leader performance.
“This predictive boost from context implies that no matter how charismatic, extraverted, or generally amazing a leader is if the situation is not conducive, then leaders will struggle.
“So leaders may want to choose their situations wisely. Just because someone offers you the opportunity to be prime minister of the UK, for instance, should you take it during these Brexit times (especially if you’re worried about longevity)? I don’t know much about Theresa May as a leader, but the cards were definitely stacked against her.”
While machine learning is influential, Spisak argues caution is necessary.
“Machine learning is ethically void. It will find patterns, but that does not mean we should act on them. Inadvertently using the wrong fuel in the machine learning engine has the potential to damage both the engine and the operator.
“Perhaps one way our future should change is the development of organisational departments for the ethical application of machine learning and artificial intelligence.”