Think Twice About Using AI to Generate Meeting Notes

While AI can generate meeting summaries, you should think twice about whether it makes sense to go down this path at your organization. Let's explore why and how to do it if you're adamant.

To resolve any confusion, I'm an AI optimist that sees how incredible the future of business is with AI involved, including all of the second to nth order effects it will have on society at large.

But, the rush to use AI to generate meeting notes, recaps, and action items is absurd.

Let's start with some of the reasons that people become fixated on this use case:

• Fear of missing important details that slip through the cracks during discussion and would be lost

• Note-taking prevents full participation because a person's time is split between contributing, listening, and drafting

• Too many meetings causes fatigue in switching between contexts and needing to stay on top of what was discussed and next steps

Those first two (missing important details and preventing full participation) are symptoms that are more directly related to one of these issues:

• Overloading the agenda to include as many topics as possible, like trying to fit a bunch of circus clowns into a tiny car

• Excessive pacing to speed through a topic, like having the Autobahn go through a school zone

• Attempting to multi-task while in a meeting to answer emails, chats, or actually produce work, like driving a car while texting and eating a bowl of soup

As for having too many meetings, it doesn't matter what the next steps are because you'll need to carve out time to do that work, and unless the next step is to schedule another meeting, you'll need to address the root cause: too many meetings.

These challenges aren't unique to your organization, but they require human solutions.

Before you throw AI into the mix, solve those underlying challenges first and consider using async collaboration to relieve some of the pressure.

Now, let's assume your organization is free of those challenges.

Are you already writing meeting notes and sharing them with your peers?

If not, start there and consider adopting a format to help with organizing key points, ideas, action items, etc.

Once you have established this within your culture, then consider whether you need to rely on AI to generate notes, recaps, and action items.

Regardless of how incredible it is for AI to simultaneously transcribe voice, organize by topics, assess sentiment, it's not going to solve for the:

• Critical thinking and synthesizing from actively documenting and reviewing others' notes

• Nuance about what should (not) be included based on how a meeting unfolds

• Neurodiversity amongst participants that you pick up on that are not piped through AI

Also, before discarding this as "old world" thinking, appreciate that you may change industries later in your career where using AI is not permissible based on privacy and security practices (e.g. investment banking, pharmaceutical research and development).

Some companies take this so seriously that they maintain their own corporate aviation fleet and have counterintelligence teams, not simply for convenience, but also to reduce exposure to competitive espionage during travel away from their offices.

TLDR: Solve any underlying challenges with your meetings, write and share your notes with peers, then consider whether AI can play a secondary role.