Everyone can agree that meetings are the most counter-productive part of their days. If people could reduce the amount of low-quality meetings they must attend, and focus on work that matters, it can go a long way to job satisfaction and productivity.
Artificial intelligence may have a new approach to this scourge — but needs to be interactive in an intelligent sense, and not just another crutch for humans caught up in onerous processes.
Consider the jobs of software developers. Brian Houck, principal productivity engineer for Microsoft, studied productivity patterns among 5,000 software engineers at his company and shared some of his findings with attendees at the Developer Productivity Engineering (DPE) conference in New York, sponsored by Gradle.
A revelation from the study was that friction in developers’ time — especially waiting for code reviews and having to spend time at meetings — caused them to lose focus. “Focused work, on the other hand, can arise out of “an uninterrupted block of time of an hour or more — no emails, no IMs, not hopping on Team calls or Zoom calls,” Houck related.
The key is “to reduce the amount of time we are in low-quality meetings,” Houck continued. “Too many meetings is the second-most cited workplace challenge for developers.” A high-quality meeting, on the other hand, would be those that involve collaboration or meetings with bosses.
“How do we tell if a meeting is a low-quality meeting?” he asked. “One way is if you’re doing something else during that meeting. If you find yourself sending emails or writing code, that isn’t you being more productive, it’s sitting in a meeting you shouldn’t be in.”
Enter AI. “Imagine if a conversational AI were listening in on the discussion,” says David Shrier, a professor at Imperial College, “automatically generating notes of what was said using speech-to-data. Imagine further that the AI was tied into out calendars and emails, so it had some ideas around the context of our conversation.”
In addition, “imagine that AI was not only able to transcribe our words, but also understand the meaning behind them, and was able to extract the major themes of our discussion, pick up verbal cues of people being assigned responsibilities, and sketch out a project plan.”
Shrier discusses these possibilities in his latest book, Welcome to AI: A Human Guide to Artificial Intelligence, in which he urges the development of human-level AI that can make everyone’s’ job not only easier, but more insightful and rewarding.
In the case of meetings, with a human-machine hybrid interaction, “we could look each other in the eye and focus on each other, with an AI in the background acting as an enabling digital assistant. Imagine what an organization could look like if AI minders were keeping track for us of what we want to remember,” he suggested.
“Imagine if we didn’t have follow-up items slip through the cracks, if we had an unobtrusively helpful machine reminding us of key workflow priorities, helping us manage our teams ad ourselves better so that a project could come in ahead of schedule and under budget,” he continued. “Further, imagine if these AI systems could assess probabilities around delivery, and could help us adjust of problems were forecast to arise.”
The possibilities, of course, go far beyond paring down meetings and extend to all aspects of work and life. The key is focusing on ambient AI that delivers human-scale benefits. “Beyond listening and understanding, beyond interpreting and forecasting, beyond even connecting, we need to bring the AI into a tighter collaboration with human systems,” said Shrier.
Chatbots, while still clunky, will help pave the way toward such collaboration. “It appears inevitable that we would take the chatbot and merge it with the calendar and other capabilities to create virtual assistants,” he adds. “While this might represent labor displacement for the human admin, it can help the busy entrepreneur or executive be more productive.”
The current thinking around virtual assistants is that they run in the background and provide guidance, but there’s a risk to their use as well. “AI assistants can make your brain a bit weak, like an underutilized muscle,” Shrier says. “This gives rise to the term ‘Google brain,’ for how people no longer retain facts and figures because they are but a few keystrokes away in the search bar.” Or think in terms of following GPS directions without really knowing what roads you are traveling.
Instead, AI needs to act as an interactive intelligent assistant. “What if we had AI that could learn what we know, and what we would like to know?” Shrier asks. “And that AI started making suggestions to us throughout our workdays to extend our thinking, perhaps even in new and unexpected directions, but in all instances aligned with our desires and goals?”
Human-AI hybrid tools need to be “appropriately aligned to what we do already, and help us do it better,” he advocates. “Where human-AI hybrids deliver superior performance are the places where we are taking the very best of what people can do and fusing the into a new creation — a ‘digital centaur’ that brings creativity and power into a seamless whole.”
Such is the mark of a truly productive partnership — without the meetings.
Read the full article here
