Descriptions:
Greg Isenberg sits down with Alex Finn for what they bill as the most comprehensive tutorial on Hermes Desktop, the recently launched desktop application for the Hermes AI agent platform. Finn, a longtime power user who previously championed OpenAI’s ChatGPT, walks through every major feature of the app—multi-session management, skills library, tool sets, cron job scheduling, and messaging integrations—explaining how each improves on the previous Telegram-based workflow.
A major theme is cost control. Finn explains that users paying upward of $1,000 per month are typically running one long-threaded conversation, inflating token costs by as much as 3–4x by sending full conversation history with every message. The session-based architecture of the desktop app lets users isolate projects and topics, dramatically trimming per-message context. He also demonstrates how to disable unused skills—Hermes ships with over 150 active by default—to cut costs further without sacrificing capability.
The latter portion covers real-world business use cases: setting up autonomous agents to handle routine tasks, using cron jobs for scheduled workflows, and managing everything through a graphical interface rather than a command-line setup. For anyone already using Hermes via Telegram or Signal, this episode is a practical migration guide that addresses the most common pain points—context bloat, opaque skill management, and CLI friction—with concrete, step-by-step solutions.
📺 Source: Greg Isenberg · Published June 06, 2026
🏷️ Format: Tutorial Demo
