“Amy” saves entrepreneur Gillian Morris about 43 productive hours a year.
Morris, the founder of Hitlist, a travel app that alerts users to cheap flights, has been using Amy, a virtual assistant from x.ai for about two years, to schedule meetings. To ask for Amy’s help, Morris sends an email to the person or people she wants to meet with and copies Amy. From there, Amy takes Morris out of the email chain and handles the back and forth about dates and times.
Morris estimates Amy schedules about 10 meetings for her a week, and spares her from having to read or respond to any related emails herself. “That’s time I can spend concentrating on my business and things that will have a better ROI,” Morris says.
Despite concerns that artificial intelligence and robots will one day replace workers–a Pew survey found that 65% of Americans expect most jobs will be automated by 2065–several companies are using AI like Amy to increase productivity. Digital agents assist with information gathering and automated tasks such as scheduling meetings or answering customer questions and requests. While these sound like small tasks, the companies that use them are reporting greater employee productivity.
Artificial intelligence is enabling customer service agents at Cable and Wireless to help customers faster and with more accuracy. In the past, Alvin Stokes, Cable and Wireless’s senior vice president for customer experience, says that customer service agents would need to have all the company’s current customer policies and promotions memorized. They would have to sift through multiple databases looking for one piece of information about the customer at a time, processing this information while on the phone with the customer.
Now, instead of having five or six screens open on their desktop at one time looking for information that might answer the customer’s question, the customer service representative works with a virtual agent to get the answers they need, Stokes says. The employee simply types or speaks the search terms, and the AI will locate and share the correct information. “It cuts down on the time it takes to get an answer and the information is more accurate,” Stokes says. It also cuts down on the stress for call center employees, especially for new hires. “New employees don’t have to memorize everything right out of the gate,” he says.
If you’ve used Amazon Echo, Siri, or Cortana, you’ve used AI.
So while some employees might feel unsure about AI, they’ve often already used it in some form but didn’t realize it, says Ed Boyd, vice president of Dell’s experience design group. He says that while many consumers are hesitant about self-driving cars, self-driving features such as rearview cameras, parking sensors, and automatic breaking have been in use for a while. “It’s just becoming more sophisticated,” he says. The same is happening with our laptops, he says.
A common misconception about AI is imagining that it’s a robot, but this type of AI doesn’t take the form of a physical machine, says Fred Brown, founder & CEO of Next IT, the company that developed the AI used at Cable & Wireless, as well as for Amtrak and Alaska Airlines. The AI lives inside different devices and is accessed through the touchpad, microphone, or keyboard.
In fact, a previous Pew survey of over 1,800 workplace experts revealed that many believe their jobs could be made easier with the assistance of artificial intelligence.
The key to AI increasing employee productivity is for the technology to learn the employee’s preferences, Brown says, not the other way around. “A human needs to train the AI so the technology does what you want it to,” Brown says. For instance, he says, if AI is being used in a call center, it’s important to listen in and review transcripts to make sure the AI is delivering a consistent experience day-to-day and week-to-week. If someone gave you a different answer each day, you wouldn’t trust it, Brown says. The same is true of virtual agents. “They need to give a consistent answer and they need to develop trust,” he says.
The U.S. Army has been using virtual agent SGT STAR to answer questions from soldiers and potential recruits for a decade, says Paul F. Denhup, chief of the experiential marketing division in the Army marketing and research group. SGT STAR has gotten better with age, Denhup says, and is now available to answer questions via goarmy.com, a downloadable app and on Facebook. “Through the algorithm and analysis of data streams, [SGT STAR’s] gotten smarter and more accurate with his answers,” Denhup says.
SGT STAR’s data also has helped the Army with its forecasting and planning, and has enabled the Army to tweak its web pages to anticipate searches and questions that soldiers and potential recruits will ask.
“The more data you feed the AI, the better it works,” Stokes says. “The more experiences the AI has, the better it can communicate and know how to handle customers in a more human way.”