Hermes + DeepSeek V4 = 100X Cheaper

Unlocking the Power of Hermes and Deepseek

Introduction to Hermes and Deepseek

• Jack introduces the capabilities of combining Hermes Agent with tools like Deepseek, emphasizing that it unlocks features unknown to most users.

• He encourages viewers to prepare for a deep dive into leveraging Hermes with Deepseek, hinting at setup guidance.

• Jack highlights the importance of understanding software's cost-effectiveness compared to traditional labor.

The Shift in Software Utilization

• The discussion centers on how AI can perform tasks traditionally done by junior developers, prompting users to consider how much time they can save using these systems.

• A distinction is made between Claude code (focused on coding within repositories) and Hermes (which learns from user interactions across various tasks).

Learning Loop and Self-Evolution

• Jack explains that every task given to Hermes contributes to its learning process, enhancing its ability to assist users over time.

• He mentions that Hermes operates continuously, even while users sleep, allowing for productivity gains through overnight work.

Model Selection and Efficiency

• The concept of selecting the best model for specific tasks is introduced; Jack recommends Claude Opus 4.7 as an effective conductor model.

• He discusses the power of Deep Seek V4 in this context, noting trade-offs when running models locally versus remotely.

Cost-Benefit Analysis of Models

• Jack emphasizes evaluating whether lower-cost models can deliver substantial value without needing maximum performance levels.

• He presents a comparison showing how much cheaper Deep Seek V4 is while still achieving similar results as more expensive models.

Integrating Open Router for Enhanced Functionality

Benefits of Open Router

• Open Router is described as a powerful tool that simplifies usage tracking and model access through a single key system.

Multi-Brain Model System

• The multi-brain model approach allows different AI models' strengths to be utilized effectively based on task requirements.

Recommended Models for Integration

• Two essential models are suggested: OpenAI's ChatGPT (for $20/month access to advanced features), and Gemini CLI (accessible via Google account).

Practical Application Using Gemini CLI

Utilizing Gemini CLI

• An example is provided where Gemini CLI analyzes video content efficiently, showcasing its multimodal capabilities.

Installation Guidance

• Instructions are shared on installing Gemini CLI from GitHub, highlighting ease of integration with various services.

Implementing the Triad Strategy

Overview of the Triad Approach

• The triad strategy involves three distinct roles: planning (Claude Opus), execution (Deep Seek), and critique (another intelligent model).

Importance of Critique in Workflow

• Emphasizing critical feedback during development processes ensures high-quality outputs before finalization.

Building Your Personalized Operating System

Creating a Dashboard with Pantheon

-Jack describes building an operating system dashboard that connects all AI tools used in workflows for better management.

Connecting Memory Systems

-He explains how integrating memory systems into Hermes enhances its contextual understanding and effectiveness.

Persona Development

-A template for creating personas within the triad system is introduced, detailing their specific functions in workflow optimization.

Engaging with Orpheus Persona

Task Management Through Orpheus

-Jack demonstrates how Orpheus manages complex tasks by coordinating input from different AI models effectively.

Finalizing Tasks with Feedback Loops

-The importance of defining desired outcomes clearly when engaging Orpheus is highlighted as crucial for successful task completion.