More The Startup Ideas Podcast episodes

Jack Dorsey's Buzz: The New Hermes Agent? thumbnail

Jack Dorsey's Buzz: The New Hermes Agent?

Published 28 Jul 2026

Duration: 00:38:47

"Buzz is an AI-first team collaboration platform that replaces Slack with native AI agents, context retention, and flexible model integration, streamlining workflows for solopreneurs and small teams."

Episode Description

I sit down with Vinny for a live tour of Buzz, an open source, agent-native chat app from Block built on an open protocol. Vinny makes the case that o...

Overview

Buzz is a new collaborative platform that integrates AI agents as core team members, distinguishing itself from tools like Slack by treating agents as first-class participants in workflows. The platform features an agentic workspace where AI can join conversations, maintain context across interactions, and switch seamlessly between different models using swappable harnesses. This design supports flexibility in model choice, reduces AI fatigue, and preserves continuity when changing systems. Buzz also enables real-time audio huddles with agents, parallel development through isolated work trees, and integration with GitHub for automated coding, deployment, and live previews.

The platform is particularly suited for solopreneurs and small teams, offering tools to streamline brainstorming, product development, and repetitive business tasks through AI automation. It supports end-to-end workflows such as generating client proposals, building analytics dashboards for social media, and deploying full-stack applications with minimal user input. Built on the open Nostr protocol and integrated with Bitcoin's Lightning Network, Buzz emphasizes data ownership, open standards, and potential future capabilities like micropayments for AI compute. Its focus on shared context, local model hosting, and open-source collaboration aims to reduce vendor lock-in and empower users with greater control over their data and tools.

What If

  • What if you built your entire solo startup workflow around AI agents as teammates?

    • Move: Set up a dedicated Buzz workspace where every project is co-developed with at least one specialized agent (e.g., a coding agent for MVPs, a copy agent for outreach). Use swappable harnesses to switch models based on task needs - like Sonnet for reasoning and Fable for creative writing.
    • Why Now?: Buzz's native agent integration and context retention eliminate the friction of juggling multiple tools; this capability is still novel and underutilized, giving early adopters a significant speed advantage.
    • Expected Upside: Reduce time-to-market for new ideas by 50 - 70% by automating ideation, prototyping, and deployment loops - all within a single persistent environment.
  • What if you replaced all manual client proposal work with AI-generated, context-aware pitches?

    • Move: Instruct your agent to ingest past client notes, brand guidelines, and industry data, then generate tailored proposals on demand (e.g., "Pitch a social strategy to an NFL team"). Use audio huddles to refine drafts in real time through voice conversation with the agent.
    • Why Now?: With Buzz preserving full context across interactions and allowing live audio collaboration, you can iterate proposals faster than text-based workflows while maintaining brand and strategic consistency.
    • Expected Upside: Cut proposal creation from hours to minutes, enabling you to scale freelance or agency offerings without hiring - turning one-person operations into high-output service engines.
  • What if you offloaded end-to-end app deployment and analytics to AI agents running on shared relays?

    • Move: Configure your agent to monitor Twitter/X analytics via API, auto-generate daily performance digests, and deploy updated dashboards to Relay-hosted environments. Connect GitHub so agents can push changes and share live previews with stakeholders without your involvement.
    • Why Now?: Buzz's relay infrastructure and GitHub integration allow agents to own full dev cycles - from data pull to deployment - while you retain control over triggers and approvals.
    • Expected Upside: Automate 80% of repetitive development and reporting tasks, freeing you to focus on high-leverage decisions while maintaining ownership of data and workflows in an open, portable system.

Takeaway

  • Set up a centralized workspace on Buzz to integrate AI agents as core collaborators for rapid prototyping and task automation.
  • Use swappable AI harnesses to match specific models (e.g., Fable, Sonnet) with tasks like coding or copywriting, optimizing performance and cost.
  • Enable audio huddles with AI agents to conduct real-time brainstorming or feedback sessions instead of relying solely on text-based workflows.
  • Connect Buzz to GitHub via relays to automate code deployment and maintain live previews of applications built or modified by AI agents.
  • Pool local compute resources with small teams to run shared AI models, reducing reliance on expensive cloud APIs while maintaining data privacy.

Recent Episodes of The Startup Ideas Podcast

5 Aug 2026 These AI Marketing Agents Get You Customers

"AI-powered marketing agents automate customer acquisition by tracking LinkedIn and email interactions, using intent-based targeting and cloud automation to enhance outreach efficiency and organic engagement."

3 Aug 2026 Graph Engineering Clearly Explained

"Graph engineering structures AI workflows into specialized, interconnected roles and tasks, improving reliability, scalability, and transparency while avoiding linear limitations."

27 Jul 2026 How I use Claude Code + MCPs to run my marketing

"Startup growth faces customer acquisition hurdles, but marketing agents - autonomous systems solving data problems - can automate and optimize these efforts, with AI-driven WordPress plugins and ad optimization as key examples."

More The Startup Ideas Podcast episodes