The Easiest AI Win at Work

The more I talk to professionals at serious organizations, the more I notice the same gap. Most teams aren't transcribing their meetings. No AI notetaker. No background transcription. No record at all beyond whatever someone remembers to jot down on their own, whether that's a Post-It, the Notes app, or a Google Doc. Nobody flags it as a problem because they don't realize it is one. I think it's one of the easiest, highest-ROI ways a knowledge worker can start using AI today, and most knowledge workers haven't gotten around to it yet.

Two examples that have stuck with me recently. The first is a friend at a major tech company, one of the most recognizable names in Silicon Valley, who told me casually that their internal meetings aren't being transcribed. The sales team probably is, because sales has been recording calls for years, but the rest of the organization? Nothing. The second is someone I know who interviewed at an M7 MBA program just a few weeks ago. Over Zoom, the interviewer was visibly typing notes. For a school running thousands of decentralized interviews in a high-stakes admissions process, there is no complete record of those conversations. Just whatever the interviewer managed to capture between questions.

These aren't resource-constrained organizations. The cost of transcribing a meeting is trivially low, and the technology runs seamlessly in the background. I'd argue 90% of internal meetings at any organization should be transcribed. Anything less is waste.

If you're looking for the single easiest, highest-ROI way to start using AI at work, this is it.

At West Point, one of the first habits drilled into every new cadet was showing up to everything with a pen and paper. Every meeting, every briefing, every field exercise. That discipline followed me into the Army and into every job after it. For a long time, being a thorough note-taker was a mark of professionalism. It was good decorum.

I think that's changing.

Here's what shifted for me when I started transcribing every meeting: I stopped splitting my brain. If you've ever been in a conversation while simultaneously trying to capture what's being said, you know the tradeoff. Part of your cognitive load goes to note-taking, and less of you is actually engaged with the person across the table.

We're heading toward a world where the professional move is full attention, because the conversation is being captured anyway. No second app open, no typing, no distraction. Just a conversation. That is actually the most respectful way to engage with someone, and it frees you to think during the meeting instead of trying to reconstruct your thoughts afterward.

If you've ever had a full day of virtual calls and found yourself drifting or attempting to multitask (and let's be honest, every knowledge worker with a packed calendar has), transcription solves for that too. When you know the record is being captured, you can actually be present.

The transcript itself is valuable. But it's only capturing what was said. It doesn't capture what you were thinking, and it doesn't capture what anyone else on the call was thinking either.

After a meeting, take two minutes to share your reflections with the AI. What you noticed, what concerns you, what you think the next move should be. When you layer your own thinking on top of the transcript, the tool gets remarkably sharp at tracking where things stand, where there are gaps in communication, and what needs to happen next.

We've all had days packed with back-to-back meetings where only the most urgent items get handled and everything else falls through the cracks until it resurfaces weeks later. A good transcription habit plus two minutes of post-call reflection is a safety net against that. You stop dropping things. You stop losing context between conversations. It compounds.

Almost every major meeting platform offers AI transcription now. Zoom, Google Meet, and Microsoft Teams all have native features, though whether they're enabled depends on how your IT team set things up. If transcription isn't available to you, ask for it. Seriously. Contact your IT department or your manager and say, "How do we get transcription enabled on our accounts?" That one ask could change how your entire team operates.

Beyond native features, standalone tools have matured fast. I started with Granola about a year ago and it was a game changer. Runs in the background, captures notes with strong accuracy, and was free for a long time. I currently use Notion's Meetings feature, which detects calls from my calendar automatically and organizes everything within my workspace. Notion also lets you build a knowledge base around those notes, which is a level up from trying to organize transcripts in Google Drive or (god forbid) a Teams folder.

There are many options. The market is competitive and getting more so as the cost of compute continues to drop. A few words of caution: if you're evaluating sales-focused tools, pay attention to pricing. Some platforms in this space are notorious for upsells once you're locked in. Do your homework before committing to an annual contract.

If you're in a regulated industry like healthcare or finance, there are compliance-specific versions. If you're at a startup, experiment freely and try several.

Recording laws vary by state, and the legal landscape hasn't caught up to AI transcription. Some states require consent from all parties before recording.

There are also real pitfalls to getting this wrong. Fortune published a piece this morning about AI notetakers creating HR nightmares, including cases where employees leave a call and the notetaker stays on, quietly recording whatever gets said next and then emailing the transcript to the full team. That's a governance failure, not a technology failure. Organizations need clear rules around when notetakers are appropriate, who gets the transcript, and how to kill the recording when conversations drift into sensitive territory.

That said, background transcription has normalized quickly among knowledge workers operating in good faith. I've almost never had an issue, and I've been doing this for over a year across a range of contexts. My practice with clients is to mention that transcription is running and offer to share the notes afterward. I'd recommend making note-sharing a standard habit. It eliminates information asymmetry, it's broadly appreciated by colleagues and clients, and I've never received negative feedback from doing it.

Where I'd draw the line is secretly transcribing and never sharing. That creates the wrong kind of information advantage. Share the notes. Everybody benefits.

More broadly, I think we're heading toward a world where most work meetings are effectively on the record, whether that's in a conference room with someone running transcription on a laptop or over Zoom with a background tool. That might be an adjustment for some, but I think it's healthy. It keeps everyone a little more focused and professional.

This gets even more powerful in a team context. When every meeting has a transcribed record, you dramatically reduce miscommunication risk. Colleagues who missed a meeting get a summary. Everyone operates from shared context instead of competing recollections. The organizations that build this habit early will have a meaningful edge.

The fact that even the biggest, most sophisticated companies haven't done this yet tells me we are very early in the adoption cycle. Which means there's an opportunity to push your organization forward, even if your company is behind. At the very least, you'll be a much better version of yourself by building this habit independently.

Transcribe every work meeting you have this week. If your platform has native transcription, turn it on. If it doesn't, try a free tool like Granola. After at least one meeting, spend two minutes giving the AI your reflections on what was discussed. Notice how it changes both the quality of the notes and the quality of your attention during the conversation. And if transcription isn't enabled at your organization, send an email or Slack message to IT asking how to get it turned on.

Austin Rief wants an AI tutor. The founder of Morning Brew posted on X this weekend looking for someone NYC-based to sit next to him and show him how people are using AI. One of the most successful media entrepreneurs of the last decade, publicly raising his hand. I may or may not have DM'd him, and I may or may not have heard back. If you’re looking for a data point on where we are in the adoption curve, that's it.

Claude with Ads. OpenAI officially started testing ads in ChatGPT today. Anthropic responded with Super Bowl ads taking shots at the decision and pledging to keep Claude ad-free. And somewhere in between, TBPN seized the moment with a parody announcement for "Claude with Ads," which Anthropic apparently did not find as funny as the rest of us did (the post was pulled from LinkedIn, still up on X).