AI In Meetings: Treading on Sacred Human Space

Intro: The Shift From "How" to "What

In the AI era, the key challenge isn’t getting things built - it’s figuring out what to build or do to begin with. More human cognitive effort will go towards ideating and planning, and less towards the ins and outs of how to bring those ideas and plans to fruition. Meetings will need to be more strategic, creative, and informed.

A Brief History: AI Everywhere

Let’s take a step back and review how we got here. When ChatGPT first took off, people usually went directly to an AI product in order to access AI. You would visit the ChatGPT website and get the answers you need. Then, as developers took advantage of APIs that let them integrate AI into their own products and services, people started seeing AI in the stuff they were already familiar with using. AI came to you.

It’s actually difficult now to find software that doesn’t have AI in it. AI is being thoughtfully added or awkwardly shoehorned into existing products, but there are also pieces of software being built from the ground up with AI capabilities in mind. AI is now being inextricably woven into some of our most essential digital experiences - Gemini is in Gmail, AI is baked into software development or website building tools, and even our To-Do lists claim to be “AI-native”.

For a while, meeting software remained untouched, sort of a rare, sacred space for human-to-human connection (albeit online) where AI was nowhere to be found. Those days are definitely over. Now, in online meetings you will see silent, note-taking AI bots from various services occupying the same squares on the grid that humans usually occupy. Just yesterday Google announced a Gemini integration into Google Meet, where it provides meeting notes. Is there nothing sacred left?

In case you think I’m lamenting the situation, let me make clear that this isn’t an ode to the good old days of meetings before AI, human gatherings in the Garden of Eden before tech companies decided that here, too, we must eat the fruit from the alluring tree of large language models.

Instead, I think AI does have a compelling role to play in online meetings. But I don’t think it will look at all how it looks now, with a bunch of disparate bots all transcribing meetings for various people.

Let’s start with why I think AI has a place in meetings.

Why AI In Meetings

Language: The Raw Material

For starters, meetings involve a lot of language, and large language models are very good at taking words and transforming and analyzing them in helpful ways. If you’re of the opinion that “AI is bad” or “AI isn’t getting any better”, then you will disagree with much if not all of the rest of this post.  

In case it wasn’t already clear, one of my core premises moving forward is that right now AI system are already incredibly good at taking the raw materials of words and doing valuable things with them. Meeting transcription and meeting notes are just the table stakes but they lay an important foundation for many other things.

We’re working on systems for large companies that will flag when different teams are discussing the same topic but clearly have different understandings of it, and it will suggest meetings that should take place between people based on their shared interests and goals. For remote companies where lots of knowledge is constructed in online meetings but often slips through the organizational cracks, this kind of company-wide transcription analysis engine is a game-changer.

There are tremendous productivity gains to be had by companies that take a “transcript by default” approach to all of their non-sensitive online meetings. This kind of radical transparency has lots of other benefits, too, but that will be the topic for another blog post.

The main point here though is that meetings are comprised of words (among many other things), and words are the raw material for sophisticated AI systems that can take that input and do a range of helpful things with them.

Creativity: Expanding The Possible

Creativity is increasingly important, and I think when leveraged effectively then  AI can help teams be more creative. AI can broaden the range of ideation, push us in unexpected directions, and help people quickly visualize or experience their ideas. Note: I’m obviously not saying it’s a hard requirement - people have always been creative and will continue to be, with or without AI in the mix.

When people come together to meet, I think there should be as little friction as possible when this question comes up: “I wonder what that would look like?”

Generative AI tools can help people prototype ideas, visualize them, revise them, and accelerate creative processes. In Frame, we’re making tools that help people use generative AI together to create images, videos, 3D models, and more, bringing things to life in real time during a meeting so that it’s a lot easier than someone screen sharing in with five tabs of various tools they are juggling.

We aren’t trying to create a new multi-modal model or be the best place to go to generate video with AI - but we are trying to be the best place for people to use generative AI together in real-time to facilitate their conversations and ideation. Once people hone in on an idea in the course of a meeting, then people or AI agents will be able to go iterate on them on their own, using the best tool for the job at hand.

More Teammates - Really Smart AI Agents

I’ve learned to approach AI with a healthy mix of skepticism and humility, and I’m happy to admit that it does most things much better than I can. In many areas it isn’t even close. I think people who aren’t finding that to be the case do not know how to effectively use the available tools. The clunkiness of the tools is sometimes to blame.

As someone who has seen how helpful AI can be across many domains, I desperately want AI to be present and accessible during meetings. When people see the results of our vision, they will want it too. In fact, I think it will seem silly not to have it!

Those who don’t want it will be the people who really want to seem like the smartest person in the room at all times. But those who are interested in results and not ego will be happy to have AI-powered teammates at their meetings.

Beyond the Bot: AI in the Next Generation of Meetings

When I talk about AI powered teammates, I’m not talking about the silent, slightly ominous note-taking bots you see jammed into Zoom calls. I’m talking about AI agents who can talk, chat, and create. Agents who might challenge you, contribute unexpected ideas, or raise evidence from other meetings that have taken place across the company. Not lurkers.

Imagine a large company with multiple teams working on overlapping product features. An AI agent tracks the discussions in real time and flags potential conflicts or synergies. It then auto-schedules a follow-up session with the right stakeholders. No more discovering after launch that two teams duplicated effort.

I’m talking about teammates who might be trained on the entirety of your company’s internal documentation, who can present research they’ve done on your competitors, and who can take on action items that you’ve set out during the meeting, even after all the people have left the meeting space. That’s what we’re building.

Conclusion: The Future of Meetings with Frame

Frames are persistent, virtual, AI-powered spaces in the browser. They aren’t video conferencing links that expire when the meeting is done. They are spatial workspaces where humans and AI agents collaborate, communicate, and create at any time.

At Frame, we’re building the future of meetings—where AI is woven seamlessly into creative and strategic collaboration. If you’re ready to transform how your team meets, ideates, and executes, join us on this journey and see how Frame can help you unlock the full potential of AI-powered meetings and teammates.