In this section I review one AI-powered application and demonstrate how it can be used to create new value.
This week the tool is Granola, an AI notepad for meetings that I have been using for the past few weeks. Its makers are insistent on one distinction: it is not a notetaker. You still show up to your meeting and jot down your own notes, the way you always have. Granola listens in the background, transcribes the conversation, and then merges your rough notes with the full transcript into something cleaner and more useful than either one alone. You stay the author. AI does the tidying up.
That framing is the whole reason I like it. Most meeting tools pull you out of the room. They join the call as a visible bot, record everything, and hand you an autogenerated summary built from the recording alone. Granola leaves you doing the one thing a human should still do in a meeting, which is paying attention and deciding what matters. It just takes over the part you were never much good at anyway, capturing every detail while also trying to stay present.
How it works
You install it, connect your calendar, and from then on it shows up in every meeting ready to go. There are no bots in the call. Granola transcribes your own computer's audio quietly in the background. In early-stage or candid discussions, the absence of that extra presence genuinely changes the tone of the call, though a quick heads-up to the other person is still good manners. It works across Google Meet, Zoom, Slack huddles, and Teams, and there is an iOS app that handles in-person meetings too.
You can also chat across all your meeting notes. I can ask, "what did we decide last time we talked about this," or "what are the open action items from this week," and it searches across everything and answers in seconds. It is less like a notes app and more like a searchable memory of every conversation I have had at work.
What makes it worth the switch
Your notes plus the transcript. Because Granola weaves what you wrote together with what was actually said, the summary reflects your judgment about what mattered. That blend is much better than tools that autogenerate a summary from the recording alone.
Recipes. These are pre-built prompts tailored to a meeting type. A sales call can trigger a coaching breakdown. A client call can generate a follow-up email draft. There is a learning curve to building your own, but once you have shaped a recipe around how you actually work, it becomes a force multiplier.
Folders and context. You can group meetings, user interviews in one place, client calls in another, and attach context documents to each. When I chat with Granola about a client, it already has the relevant background, so the answers are sharper.
Granola helps you stay present in your meetings and makes sure nothing falls through the cracks afterward. Most tools make you choose between being in the conversation and capturing it. Granola lets you do both, and that is what has kept it on my machine these past few weeks.
It also fits a theme running through this issue. Granola earns its place through how it splits the work: the AI handles the tidying, and you keep the judgment about what mattered. That is the same human-led lever I write about in this issue's piece on speaking the language of your audience.
Your action step
Pick one recurring meeting you always leave with messy notes, a weekly team sync or a client call, and run Granola alongside it for two weeks. Keep taking notes the way you normally do, and let it merge them with the transcript. At the end of the fortnight, ask it across all those sessions: "what did we commit to, and what is still open?" If the answer saves you the scramble of reconstructing it yourself, you have found where it fits.
You can try the free plan yourself at granola.ai, and as always, reach out if you would like help thinking through where a tool like this fits in your workflow.