For years, I took notes in meetings. Then I started using AI to transcribe everything. Here’s what changed.
The Setup
I use a simple workflow:
- Record the meeting (with permission)
- Upload to an AI transcription service
- Ask the AI to extract action items and key decisions
- Share the summary with attendees
What I Learned
1. I Was Missing 40% of What Was Said
When I compared my notes to transcripts, I was capturing less than half the content. Not because I wasn’t paying attention, but because:
- I can’t write that fast
- I was filtering what I thought was “important”
- I missed context and nuance
2. People Say More Than They Realize
Transcripts reveal:
- Off-hand comments that become important later
- Concerns that weren’t fully articulated
- Ideas that got lost in the conversation
3. Action Items Became Unambiguous
Before: “Follow up on the thing we discussed”
After: “Email Sarah by Friday with the Q2 budget proposal including the revised marketing spend numbers”
4. I Stopped Taking Notes and Started Listening
This was unexpected. Without the pressure of note-taking, I:
- Asked better questions
- Picked up on body language
- Engaged more naturally
The AI Prompts I Use
Extract all action items from this transcript. For each:
- Who owns it
- What's the deadline (if mentioned)
- What's the specific deliverable
Summarize the key decisions made in this meeting. Include:
- What was decided
- Who agreed
- Any dissenting opinions
List any open questions that weren't resolved in this meeting.
The Downsides
- Storage - Transcripts add up
- Privacy - Need to be careful with sensitive discussions
- Dependency - I’m worse at note-taking now
Verdict
Worth it. The time saved on note-taking and the clarity on action items alone justify it.