We’re a week and a half away from Charter’s first AI-focused summit, Leading with AI. As we were putting together the agenda for the event, several topics stood out as essential for the day, including how to train your workers in AI.
About a quarter of workers globally say they’ve received training on how to use AI for their work work, according to a survey by the Adecco Group. Three-quarters of workers report a lack of confidence in their ability to use the technology. What are the best ways to design AI training and encourage employees to experiment with it in their work? That will be the focus of one of our panels with Trena Minudri, chief learning officer and VP of talent management at Coursera, and Rebecca Hinds, head of the Work Innovation Lab at Asana.
Hinds’ team at Asana recently spent two weeks analyzing the AI adoption patterns of the company’s customers. As a preview of what’s to come at our summit, we spoke with her about some of the factors that influence whether or not someone adopts AI. We also asked her about recent research into the impact of AI on collaboration and work satisfaction. Here are highlights of that conversation, edited for length and clarity:
You find that people who are already delegators are more likely to adopt AI. Does that mean that managers tend to assign work to AI more than individual contributors?
Because Asana is a work management platform, one of the primary activities it’s used for is task delegation. So we’re able to understand at baseline who is more likely to assign tasks to other people. We were then able to see who is most likely to adopt AI. It turns out it’s the people who had already built that muscle for delegating work. We do see higher adoption within managers and executives compared to individual contributors.
Every organization needs to be paying really close attention in terms of who’s adopting AI within their organization and for which purpose. We know certain people—whether that’s role, function, age, gender—are more likely to adopt AI. Training has to be tailored to these natural propensities and biases we know certain groups have to adopt the technology.
You also found that there’s this social contagion aspect to AI adoption. When people you’re collaborating with use it, you’re more likely to use it too. What are the most promising channels for taking advantage of that social contagion effect?
One of the most powerful drivers of adoption is at the functional level. Where I see AI training break down super quickly and adoption fall flat is when the training tries to be too generic such that it’s not tailored to the specific role. When marketers see marketers using AI, there’s a very powerful contagion effect because you can see yourself in that use case.
When [training] tends to be broader across multiple different use cases, people struggle to apply it to their individual job. There’s also a homophily effect associated with all of this. The more you can make training unique to the individual, whether that’s job function, whether it’s level within the organization, whether it’s gender, whether it’s age, it’s also going to impact your willingness and your confidence in the technology that ultimately will drive you to adopt.
Some recent research suggests that AI could lead us to collaborate less and reduce our satisfaction with work. What do you make of those concerns?
We need to be careful about broad stroke conclusions about AI, especially when you have a study that’s focused on such technical workers. We know that the impacts are not going to be the same on other workers. There’s also a sweet spot. If you can apply AI to automate the monotonous parts of your work, then the most successful people are going to take that time savings and apply it to the areas of work that shouldn’t be efficient: Developing relationships with one another and being creative.
Yes, if we automate all of our work, we’re going to have less job satisfaction and we’re going to feel less connected. We know that loneliness in the workplace is a massive problem. The most successful people and teams will figure out what are those parts of work that I should delegate to AI and what does that enable me to do in terms of making the other parts of my work even more human?
Do you think AI could improve the way we collaborate?
Absolutely—I’m probably most excited about the opportunity for AI to help us become much more data-driven about how we do teamwork. We’re seeing some interesting research on how AI is going to transform organizational structures. I definitely foresee a world quite quickly in which teams can be formed, not just based on what the org chart says, but on the exact right mix of personalities, of skillsets, of prior work experience, of diversity, to optimize outputs.
Imagine launching a new product or working on any initiative and being able to look across your entire employee base to find the exact right composition of people who are going to work well together and drive that project forward to success. It has enormous potential to help us improve how we collaborate in a way that is much more data-driven and less reliant on chance.
—
Sign up here to join us on January 30 in-person or virtually for our Leading with AI Summit, where Hinds and our other expert speakers will share many additional tips for working and managing more effectively using AI.