Iif you ask chat stuffcommon technologyAn artificial intelligence (artificial intelligence) tools are all the rage, and the responses you get are almost instantaneous, completely deterministic, and often wrong. It’s a bit like talking to an economist. Questions raised by technologies like Chatcommon technology Generate more preliminary answers. But they are questions managers should start asking.
Your browser does not support
One issue is how to address employee concerns about job security. It’s natural to worry.one artificial intelligence Making your expense processing easier is one thing; one artificial intelligence People prefer to sit next to each other at dinner parties.Know how workers will reallocate the time and energy freed up by work artificial intelligence Helps promote acceptance.The same goes for creating a sense of agency: by MIT sloan management review The Boston Consulting Group found that the ability to overthrow a artificial intelligence Make employees more likely to use it.
Do people really need to understand what’s going on inside artificial intelligence not sure. Intuitively, being able to follow an algorithm’s reasoning should trump not being able to. But a study by academics at Harvard, MIT, and Politecnico di Milano suggests that too much interpretation can become a problem.
Employees at luxury brand portfolio Tapestry have access to predictive models that tell them how to allocate inventory to stores. Some use a model that explains its logic; others use a model that’s more of a black box. It turns out that workers are more likely to reject models they can understand because they mistakenly trust their intuition. Workers, however, are willing to accept decisions about a model they don’t understand because they trust the expertise of the people who built it.Credentials of the man behind the scenes artificial intelligence matter.
The way people respond differently to humans and algorithms is an emerging field of research. In a recent paper, Gizem Yalcin of the University of Texas at Austin and her co-authors examined whether consumers respond differently to decisions when they are made — for example, approving someone for a loan, Or join a country club. machine or man. They found that people reacted the same way when they were rejected. But when they’re approved by an algorithm rather than a human, they feel less positive about the organization. reason? People are good at explaining unfavorable decisions, no matter who made them. When evaluated by machines, they have a hard time attributing successful apps to their own charming, delightful selves. People want to feel special, not be reduced to a data point.
Meanwhile, in a forthcoming paper, Arthur Jago of the University of Washington and Glenn Carroll of the Stanford Graduate School of Business examine people’s willingness to give rather than take credit—particularly for those There is no work done on your own. They showed volunteers something that belonged to a particular person — such as artwork or a business plan — and then revealed that it was created with the help of an algorithm or with the help of a human assistant. Everyone trusted producers less when they were told they were being helped, but the effect was more pronounced for jobs involving human assistants. Not only did participants see the job of overseeing an algorithm as more demanding than overseeing a human, but they also felt it was unfair for someone to take credit for someone else’s work.
Another paper by Anuj Kapoor of the Indian Institute of Management Ahmedabad and his co-authors examines whether artificial intelligences or humans are more effective at helping people lose weight.The authors studied weight loss achieved by subscribers to mobile apps in India, some of whom used only artificial intelligence Coaches, some of whom also used human coaches. They found that people who used a real trainer lost more weight, set themselves stricter goals, and were more picky about recording their activity. But those with a higher BMI did not perform as well as their lighter counterparts under human coaching. The authors speculate that heavier people may be more awkward when interacting with others.
The picture that emerges from such studies is muddled. It is also dynamic: just as technology evolves, so do attitudes. But one thing is very clear.chat impactcommon technology and others artificial intelligenceIt’s not just about what they can do, but how they make people feel.■
Read more from Bartleby, our columnist on management and work:
The Curse of the Corporate Headshot (January 26)
Why Blaming Doesn’t Help (January 19)
How to Unleash Creativity in the Workplace (January 12)
To stay on top of the most important news in business and technology, subscribe to The Bottom Line, our weekly newsletter for subscribers.
