Artificial intelligence (AI) has caused a flurry of excitement—and worry. There is concern about privacy breaches, worry that it won’t be as “intelligent” as promised, that we’ll become reliant on it nevertheless, and that terrible problems will happen as a result of that reliance.
All that said, I still think it could be helpful. In particular, I’m dreaming of an AI executive assistant for all of us. At my company, not even the CEO of our division has an executive assistant all to himself. He has one, but one he shares with a handful of other executives. And people like me—and like the vast majority of mid-level employees—don’t have any.
What if we each had our own Siri or Alexa—or Howard (why are these digital assistants always women?)—who could seamlessly export data from our work computer and then do what a secretary from the mid-20th century would do: remind us of upcoming appointments, appointments that need to be made, and even birthday presents we need to buy? How about if it even remembered potential sticky situations: “Remember, you don’t like him, so be sure to get there late enough to know where he’s sitting, so you can sit elsewhere”?
You can do that now by setting up alerts on your phone, computer, or other digital devices, and sometimes a smartphone will anticipate your patterns and ask you if you want an alarm set, or whether there is another “suggestion” it should act on, but those alerts often require the user to set up notifications and input information into calendars. You can get a digital assistant such as Siri or Alexa to verbalize your alerts, but those digital systems are not necessarily secure enough for work-related material.
The AI systems I’m thinking of would be as secure as a work computer that a company’s IT system has hopefully set up security protocols and firewalls to protect, and enable the user to interact with it on nearly as casual a basis as she would with a live person who is an executive assistant. The user would be able to simply say aloud, “John, cancel my 2 p.m. appointment today and call Lucy Simmons to see if she can reschedule our meeting for next week. Be sure to apologize to her, so she knows she’s still a priority for our company.” The user wouldn’t have to give John, the digital assistant, a script for him to dictate; it would be “intelligent” enough, like a good human executive assistant, to figure out the right verbiage himself.
I saw an article in Computerworld by Matthew Finnegan last week highlighting news from London’s Digital Workplace Summit. Experts from Gartner are predicting AI at work will be commonplace by 2025. Gartner expects AI tools to help support managers in making better decisions. “Robo-bosses [will] become common in 2025. We are not necessarily saying that everyone is going to be reporting to an algorithm, so you can breathe a little bit easier,” Finnegan reports Matthew Cain, vice president and distinguished analyst at Gartner, as saying.
So my dream of a digital executive assistant may not come to fruition, but another dream—of possibly superseding toxic bosses—may. It sounds to me like a great digital boss may end up overriding an inefficient boss who, in addition to being unpleasant, is woefully unprepared and disorganized (yet bountifully arrogant). The digital boss will be super-organized, and will remember exactly what the human boss said and did, and will spit that information back for employees, and the boss herself, so there may be less “I didn’t say that” or “You must have misunderstood.” Finnegan reports Cain as putting it this way: “Your manager won’t be replaced by an algorithm, but your manager will be using a lot of AI constructs to help improve and to make more efficient a lot of the routine work that they do. We think that is going to be the combination.”
AI also may help companies optimize space, and make climate control—the much-fought-about office air temperature—consistent. Maybe it will be able to field verbally spoken complaints or e-mailed or text messages commenting on the comfort or discomfort of the air temperature and then calculate the optimum temperature based on that feedback. It could send text messages to all the people working in the office on a given day: “How’s the office feeling today? Too hot? Too cold?” When the decision-makers—the ones who literally and figuratively set the temperature for the organization—have objective data to rely on, which ideally everyone can see, it’s hard to insist it just happens to be perfect the way they, personally, like it.
Are AI systems utilized in your office yet to support employees? Do you see the current systems, and future systems, potentially making your workforce more efficient and happier?