Stop prompting your AI agents. Start managing them.
The speaker discusses the shift from traditional agent prompting to agent management, highlighting the limitations of local Kanban board approaches and advocating for cloud-based VPS solutions with multiple communication channels to effectively manage autonomous AI agents across projects.
Summary
The transcript opens with an observation that while many people discuss orchestrating multiple agents autonomously, few actually implement it successfully. Most implementations still rely heavily on human-in-the-loop prompting, even when agents are organized in loops or hierarchies with sub-agents.
The speaker then transitions to their personal insight: moving from the role of 'agent prompter' to 'agent manager' represents a fundamental shift in approach. They describe initial attempts using Kanban boards as a management tool, which seemed promising for task initiation but proved problematic in practice. The speaker identifies two specific challenges: difficulty maintaining multi-turn interactions (beyond 2-3 turns) and the limitations of running agents on local infrastructure.
The solution the speaker implemented involved two key changes: migrating from local runtime to cloud-based VPS hosting and establishing multiple communication channels to interact with agents. This infrastructure change enables flexibility in how agents are prompted and managed—through text messages, integration with Linear (a project management tool), or direct shell prompts. This multi-channel approach represents a more sophisticated agent management paradigm compared to single-interface solutions.
Key Insights
- Despite widespread discussion of autonomous multi-agent orchestration, most implementations still rely heavily on human-in-the-loop prompting rather than true autonomy
- Kanban boards are insufficient for managing agent workflows because they struggle to maintain interactions beyond 2-3 turns and lack proper intervention capabilities
- Local runtime environments do not work effectively for agent management, requiring migration to cloud-based infrastructure
- Multi-channel communication interfaces (text, Linear integration, shell prompts) enable more flexible and effective agent management than single-interface approaches
- The conceptual shift from 'agent prompter' to 'agent manager' represents a fundamental change in how practitioners approach autonomous agent systems
Topics
Transcript
[0:00] We have heard a lot of people talk about orchestrating many agents autonomously across a project. We actually haven't seen many people do it. I still see a lot of prompting even if you're prompting into a loop [music] or a goal as far as sub-agents. People are really still human in the loop. >> What really clicked for me was starting to move away from being agent prompter to agent manager. The first thing that everybody tried was a Kanban board. You would put all these things in there and move them back and forth. What I found is that it was hard to get two, three, four turns. Through that, it was easy to [0:31] get to…
Full transcript available for MurmurCast members
Sign Up to AccessMore from How I AI
AI picks my thumbnails now
A content creator demonstrates using OpenAI's Vision API to automatically analyze video frames and select the most flattering ones for YouTube thumbnails. The frame picker app scores frames on expression, clarity, composition, and impact, then combines selected frames with DALL-E 3 to generate polished thumbnail designs.
Codex built my weekly Spotify playlist from Reddit
A developer created an automated website that solves shared Spotify access issues by building personalized weekly playlists from Reddit's music communities. The system scrapes popular music recommendations from Reddit, ranks them by voting, and automatically generates a fresh playlist every Monday morning.
How the OpenAI team uses ChatGPT Sites daily
Kat, a website product leader at OpenAI, demonstrates how the ChatGPT Sites feature with plugin connectors enables diverse use cases ranging from incident management and team collaboration to creative projects like music discovery and 3D game development. The discussion highlights how Sites serves as both a practical business tool and an infrastructure platform for rapid prototyping and creative expression.
Jev beat an LLM at blitz chess
Jev, an AI system, defeats an LLM at blitz chess by using a two-step analysis process that evaluates the top three moves and their subsequent branches in under a second. While LLMs could eventually solve the same problem, they would require significantly more computational resources and time, making Jev's specialized approach more efficient for time-constrained decision-making tasks.
Jev mapped voice to color over the weekend
Jev built a real-time application that maps voice input to colors and emotions using OpenAI's real-time voice API, Java, and a quotes API. The system analyzes emotional tone and returns corresponding colors and relevant quotes in real time.