By Elise Keith
“My goal is to make sure no one ever has to write meeting notes again.” I’m talking with a product manager working to perfect automated call transcription. If find idea both exciting and problematic.
Fully automated transcription is just the first step towards a grander vision.
As an AI enthusiast told me, “Notes are just the start. Consider – what if AI could make sure the meeting goes well in the first place? What if an AI agent kept the discussion on track? What if it knew how to help a group resolve a conflict? That day may not be too far off.”
He’s painting a vision of a comprehensive augmented meeting intelligence agent.
“Augmented intelligence (AI)… is a complement—not a replacement—to human intelligence. It’s about helping humans become faster and smarter at the tasks they’re performing.”
The idea that we might soon have an AI agent operating as a virtual personal assistant, communications coach, and expert facilitator all-in-one gets people excited.
After all, meeting well requires a lot of work! Estimates show upwards of 55 million meetings in the U.S. everyday; so many tasks for AI to help us perform faster and smarter!
Today’s Reality: It’s Time to Get Excited
There aren’t any all-in-one solutions yet, but there are technologies addressing parts of the problem.
Automated Transcription and Conversational Intelligence
Real-time natural language processing and sentiment analysis software works to solve the “meeting notes problem,” automatically generating text from audio files then searching it for juicy insights.
Machine transcription isn’t perfect, but it gets enough right to catch the main ideas. Then you can analyze the records.
What topics come up? How’s the engagement? What do successful people do in their meetings to get results?
This last question drives much of the current development. Many AI products target sales and support meetings.
Why? First, managers can easily distinguish a successful result from a fail here giving customers a reason to buy AI technologies despite their imperfections.
Gong.io, for example, doesn’t market meeting notes. They market their ability to win more sales.
Sales and support meetings are also short and often involve just two people. If you’ve ever dictated audio to your phone, you know that while the technology is getting better, short and simple work best.
AI technologies can provide important clues about how the meeting went. You can’t count on AI to provide all the answers, though, so MeetingQuality combines AI insights with human reviews to predict sales and project success.
” We compare the tone detected by AI from the salesperson (mostly joy) to that of the customer (mostly sadness),” shared Kelvin McGrath, MeetingQuality’s CEO. “Comparing emotions from the rep and the customer, to the next steps, and the customer’s rating of the meeting, provides serious predictive capability on the likelihood of success.
One surprising finding from project meetings: whenever stakeholders exhibit disgust, the project is almost irretrievably lost no matter what other project indicators report.”
Yikes! How can AI prevent this kind of failure?
AI technologies are pretty good at detecting interaction patterns (pauses, trigger words, etc.) and making a real-time assessment of the meeting’s emotional tone.
Cogito uses this to deliver conversational coaching tips to support agents.
Meeting with a frustrated customer, an agent might see, “Frequent overlaps: Please pause and listen.”
Because it’s AI and not human, agents get “conversational collision detection” cues on every call. That’s the promise of AI – it never gets bored or tired. The cues are simple, but what they lack in sophistication they make up for in scale.
AI can handle an increasing number of simple tasks, because it’s always learning. This learning requires lots of data and expensive training to recognize each customer’s specialized vocabulary – like industry lingo and product names.
This means the benefits of today’s AI meeting assistants can best be realized in high-volume sales and service organizations.
That just leaves the other 90%+ of meetings!
Augmenting Technology for More Complex Business Meetings
Most meetings involve more than two people in a variety of complex situations. It’s far from clear when or even if there will ever be a robot facilitator sophisticated enough to help these meetings succeed.
However, there are existing technologies can augment the intelligence you bring to more complex meeting types.
Smart Technology Solves Specific Meeting Problems
A quick search for software designed to support the specific kind of meeting you’re running will reveal many more examples.
Augmented Intelligence Means We’re Still Responsible for Results
So why do I find the prospect of automated notes problematic?
Meetings are by definition a gathering of people seeking a mutual understanding of their common future.
What matters is the understanding created in the minds of the people, not the mind of the machine. How do the people feel afterwards? More importantly, what are the people going to do next?
Often this value crystallizes only when people take the time to discuss and write it down.
Ultimately it doesn’t matter what the robot observes. It’s what people take away that gives a meeting impact.
Will AI technology deliver us magical meetings?
No. Meeting success depends on human interaction and interpretation – technology can’t “deliver” that in the same way you deliver a pizza.
I believe, however, that when used to reveal ways we humans can improve, technology will dramatically augment our ability to make meeting magic.
The good news: once you get specific about the kind of meeting, today’s technology can put that magic within reach.
J. Elise Keith is the co-founder of Lucid Meetings and the author of Where the Action Is: The Meetings That Make or Break Your Organization. For more information, please visit, www.lucidmeetings.com and connect with her on Twitter, @EliseID8.