How can you encourage managers to use AI for decision making?

Some managers may be reluctant to use artificial intelligence as a tool for decision making even though there is a lot to gain from doing so, one researcher claims.

The algorithms is the main source of that reluctance, says, one of the authors of the report titled Exploring the Individual Factors Influencing Managers’ Attitudes Towards the Use of Artificial Intelligence in Organizational Decision-Making, from the ESCP Europe Business School. The author wished to remain anonymous.

“The main reason why people don't want to use [AI] is because they are not familiar with algorithms,” says the source in talking with Canadian HR Reporter.

“So there's a lack of understanding of what is going on. Because most of the time, algorithms are sort of a black box... And that's why they don't really trust in the algorithm.”

People are also not willing to let an error from an algorithm slide, she says.

“If an algorithm makes a mistake, people are not really forgiving. They think: ‘Okay, the algorithm will make the mistake again’. When you make some mistakes, people think: ‘Okay, making mistakes is human’. So they are more willing to forgive and trust a person again.”

Overall, 39 per cent of workers globally are feeling overwhelmed by the amount of change AI may bring to their jobs in the future, according to a previous report.

However, the source explains that the aversion against the use of AI is far more common among older workers.

“If people are more senior in their jobs, they're also less willing to use [AI]... People that did the same job for 15 years think: ‘Okay, I really know how all of this works. And I don't really need an algorithm, I can make the decisions myself’.

“And if someone just entered the company, they are more willing to use algorithms to help them with decisions. And also, younger people, they grew up with technology, so they are more used to it.”

Pros and cons of AI in decision making

In her report, the source explains that there is a lot to gain if managers would be willing to use AI for decision making. These include:

• Process Optimization and Improved Decision-Making. Algorithms can uniformly and consistently go over a vast number of alternatives. 
• Improved Efficiency. Algorithms can accelerate the decision-making process, particularly important in high-velocity contexts.
• Removing Cognitive Bias. Algorithms can be used to remove cognitive bias.
• Replicability of Outcomes. Algorithms use standard, uniform, computational processes to come to a decision.

She admits, however, that there are also cons in adaption:

• Taking on Bias. Algorithms can take on bias.
• Specificity of the Decision Search Space. The decision space of an algorithm needs to be clearly specified and restricted.
• Interpretability of the Process. Difficult to understand how an algorithm came to its decision.

Incorporating the Artificial Intelligence and Data Act (AIDA) into the federal government’s proposed Bill C-27 is simply not the way to go around legislating AI, according to one expert.

How to encourage managers to use AI?

Still, AI can be useful in the decision making process in these areas, according to Upwork:

• Predictive analytics
• Risk assessment and mitigation
• Decision support systems
• Recommender systems
• Optimization and resource allocation
• Fraud detection and prevention

So how can employers encourage their managers to use AI in decision making? Start by explaining the gains to them, says the source.

“If you explain the benefits of AI, people are more willing to use it. So if you say, for example, that AI can [work for making] different decisions because they collect a lot of data and go over all the data which humans scan, then it's way easier to make a decision involving all the data.

“Also, if you explain how the algorithm works, and [you are] transparent even if they have shortcomings, people are also more willing to use it… If you explain to them how it works, and when managers find it logical, they're also more willing to use it.”

However, in no way should AI replace humans in the decision making process, she says.

“It's good to have a good balance between automated and human made decisions. So managers don't want algorithms to replace them fully, but more like to help them. 

“So for example, algorithms can give them [the] top five best choices, and then humans can still make the decision. So [they get] help from the algorithm and they can still make a decision themselves.”

The source also notes that it’s also important to involve the managers in the design process of the algorithm “so they also really understand what is going on”.

Thirty per cent of Canadian companies are regularly employing AI platform ChatGPT to improve operations, compared to two-thirds (65 per cent) of U.S. companies, according to a previous research by KPMG in Canada.