Fireflies Bets That Dictation Is the New Meeting Note

Fireflies Bets That Dictation Is the New Meeting Note

Fireflies.ai launched "Talk" on September 29 — free, unlimited voice dictation built into its desktop app for Mac and Windows, available on every plan including free [4][5]. Hold Fn or Ctrl+Win and it transcribes, strips filler words, adds punctuation, and supports 90+ languages across any app, not just meetings. It even auto-generates @mentions when dictating into Slack.

CEO Krish Ramineni's pitch is simple: "People speak roughly 3x faster than they type... Talk takes Fireflies beyond meetings." No audio is retained — transcripts are stored locally — positioning this as a trust-forward move while Fireflies expands toward a broader "AI teammate" vision that includes voice agents [6].

X commentary has drawn quick comparisons to standalone dictation tools like Wispr, Willow, and Superwhisper, with several threads debating whether hands-free, real-time transcription is the next default input method for knowledge work generally — not just calls.

Open-Source Knowledge Graphs Challenge Paid PKM Tools

A cluster of MIT-licensed repos has landed recently, led by the Smart Second Brain plugin for Obsidian, which adds semantic search, interactive knowledge graphs built from links/tags/content, and AI agents that can read and write notes, charts, and dashboards — all running locally [7]. Companion projects like brain-map turn any folder of markdown notes into a zero-dependency interactive graph, and open-second-brain adds local-first memory for agent frameworks [8][9].

Together they read as a direct challenge to paid personal knowledge management products: free, self-hosted, and MCP-compatible, with no cloud lock-in. For anyone already sitting on a pile of meeting transcripts, the appeal is obvious — turn that archive into a navigable graph without a subscription.

X threads are picking up on exactly this use case: builders discussing how to feed meeting transcripts and voice notes into these open graphs for AI retrieval, essentially DIY-ing the knowledge base layer that commercial tools charge for.

Verbels Pushes Meeting Transcripts Straight Into Jira and Asana

Verbels is a newer entrant automating the last mile of meeting follow-through: it joins Zoom, Meet, or Teams calls automatically, transcribes the conversation, extracts action items with owners and deadlines, and — after human approval — pushes them directly into Jira or Asana [10][11]. It's currently in waitlist mode for early access.

The pitch is narrower than Granola or Fireflies: rather than building a broader knowledge base, Verbels is laser-focused on closing the gap between "we discussed it" and "it's a tracked task." Early buzz is modest but pointed — mostly from teams interested in killing manual task-entry after calls.

What This Means For Your Meetings

The throughline across today's news is consolidation of control and expansion of surface area. Granola is locking its data behind an official API while still letting AI agents in through a sanctioned door. Fireflies is expanding capture beyond the meeting itself into everyday dictation. Open-source projects are proving that knowledge graphs don't require a subscription if you're willing to self-host. And Verbels shows the market still has appetite for turning talk into tracked work, automatically.

For professionals, the lesson is the same one Proudfrog has been building around: the value isn't in the transcript, it's in what you can retrieve from it later, and who controls that retrieval layer. Whether that's an MCP connector, a dictation shortcut, or a local graph, the winners will be tools that let you ask your own meeting history a question and get a sourced, trustworthy answer — not just a searchable archive you have to dig through yourself.

Key takeaway: Meeting intelligence is splitting into two camps — platforms gating access behind official APIs and paid tiers, and open-source tools giving you full control locally. Choose based on how much you need your knowledge base to be queryable by AI versus owned outright.

Sources

1. https://www.granola.ai/blog/granola-mcp
2. https://openai.com/business/plugins/granola/
3. https://superframeworks.com/articles/granola-encrypted-local-notes
4. https://fireflies.ai/blog/introducing-fireflies-talk
5. https://techcrunch.com/2026/09/29/fireflies-adds-dictation-to-its-desktop-notetaking-apps/
6. https://www.theweek.in/news/sci-tech/2026/10/05/firefliesai-can-ai-now-make-dictation-work-across-every-app.html
7. https://github.com/your-papa/obsidian-smart2brain
8. https://github.com/zubair-trabzada/brain-map
9. https://github.com/itechmeat/open-second-brain
10. https://ng.linkedin.com/company/verbels-ai
11. https://verbels.app/