From platforms that make product recommendations based on past purchases, to chat bots that answer basic questions, artificial intelligence (A.I.) applications have made their mark on consumer experience in recent years.
And now, they’re quickly advancing in the workplace, too.
A new study by Oracle and Future Workplace found that half (50%) of workers use some form of A.I. at work, up from 32% last year. By 2021, 25% of digital workers will use a virtual employee assistant (VEA) daily, Gartner predicts, up from less than 2% in 2019.
But enterprise A.I. adopters would do wise to take a note from firms working in consumer-facing A.I. already, as they’ve have had time to work out some of the technology’s kinks. Here are four lessons businesses can learn from watching consumer A.I.
Choose the right channels
Just because you have this nifty new tool at your disposal doesn’t necessarily mean that it should be deployed everywhere, says Ryan Duguid, chief evangelist at Bellevue, Washington workflow automation firm Nintex. Duguid cautions that companies should make deliberate decisions about where the technology can best work for their employees.
“Does your solution actually make life any easier, reduce [the employees’] cognitive load, get things done quicker, or whatever it happens to be?” he asks. If not, you may want to forego an A.I. solution.
Case in point: Chat bots. While they can be a welcome facilitator when a situation calls for a quick, easy answer—like when consumers have a question about a product, or if an employee is looking for an open room to hold a meeting—they definitely have their limits, says Erik Brown, a senior director at West Monroe Partners a Chicago-based technology and management consulting firm.
“They’re not going to be effective when you’re booking a complex travel itinerary where you need to figure out the different options you have for logistics and timing, schedule a rental car and hotel, and provide credit card, et cetera,” Brown says. So beware of layering in more technology than your use-case can handle, because it could end up frustrating your employees and leaving them to clean up the mess.
Watch the data
A.I. is, of course, only as good as the data on which it is based. If companies don’t account for shifts in data or monitoring for bias, they’re going to run into trouble with A.I.’s outcomes, Brown says.
For example, West Monroe is working with utility companies to predict fluctuations in power demands. But as more solar and wind power contribute to energy supplies—or when storms become a disruptive force—relying solely on that data to deploy personnel or other resources is a risky proposition, Brown says. For all its informed decisions, one thing A.I. lacks is a good set of eyes.
“We’re going to have to be smart about continually training and retraining data as we look to execute more models,” he says.
Pay attention to security and privacy
High-profile data breaches like those at Equifax and Target have illustrated the costly and damaging consequences of mishandling data or insufficient data security measures. Privacy and security should be top of mind when handling employee data too.
Take Denver-based software development company Itransition the team developed a facial recognition technology “for fun,” says Igor Efremov, head of recruitment. It was so effective that they implemented it as part of the firm’s security system, using facial recognition along with pass cards as second-factor authentication.
But, as the firm has looked for other applications, concerns have arisen. “Definitely it does carry some implications that are not too positive and maybe some even negative,” Efremov says. Keeping employee data secure is one issue.
And A.I. issues may run afoul of more than your employees. Depending on the application, privacy violations could run afoul of the EU General Data Protection Regulation (GDPR), the U.S. Health Insurance Portability and Accountability Act of 1996 (HIPPA), and other laws.
Regulation related to consumer data control remains in the spotlight. California has a 2020 ballot measure that would give consumers more control over their data, and Democratic presidential candidate Andrew Yang has made data control part of his platform, So, companies will need to be mindful of how that affects the ways in which the employee data they collect, including through A.I. platforms, is used.
Put people first
While it may be tempting to let A.I. more functions, end-user experience should drive the decision-making. Duguid points to e-commerce shoe and clothing seller Zappos, which is widely recognized for its exceptional customer service team.
“They’re just going to get you talking to a person straight away,” Duguid says. “They know that person can help you solve your problem, in a shorter period of time. That person is empowered to make decisions about, how to make your life better, how to refund things.”
Humans still need to oversee A.I., Brown says. While the technology is improving, many functions, especially those with higher degrees of complexity, are still going to need human intervention. Ensuring that outcomes aren’t biased, using soft skills to defuse and solve situations, and ensuring that employees get the help and resources they need are essential to employee experience, but not yet things that AI-powered systems can offer without a human helping hand.
“There’s all this discussion around all of these jobs being replaced by AI and robots in the future. I think the jobs that are redundant and repetitive for humans are the ones that are going to be replaced, and that’s going to allow humans to interact with AI, interact with robots in some cases, and allow them to focus on problems that they need to solve that the human touch has to be a part of,” Brown says.
By Gwen Moran
Authors propose an iterative cycle of activity that helps companies optimize for both business value and social value in a single business model.
90 Day Finn programme is an intriguing example of how smaller hubs can think outside the box to compete with the San Franciscos, Londons and Singapores of the world.
The lean startup approach for developing new products or businesses has become popular among companies large and small. But is the approach really effective?