46 lines
1.6 KiB
Markdown
46 lines
1.6 KiB
Markdown
progress:: {{renderer :todomaster}}
|
|
icon:: 📅
|
|
boox-notes:: 
|
|
id:: 1fc8ffcf-8fa6-4a7e-8982-89adc0977d09
|
|
|
|
- ### Tasks
|
|
-
|
|
- ### Notes
|
|
- [[KaraKeep-Highlights-Summary]]
|
|
- {{renderer :mermaid_00f4ca0a-1de7-44e6-b203-d72ee1a887ea, 3}}
|
|
- ```mermaid
|
|
mindmap
|
|
root((AI Agent Systems & Deployment))
|
|
Multi-Agent Systems & Handoffs
|
|
Explicit Prompted Instructions
|
|
Conditional Edges & Command Objects
|
|
Fragile Implementations
|
|
Tension: Flexibility vs Scripted Control
|
|
Agent Framework Ecosystems
|
|
LangChain
|
|
High-Level Rapid Assembly
|
|
Provider Agnostic Design
|
|
Middleware for Customization
|
|
LangGraph
|
|
Low-Level Fine-Grained Control
|
|
Production-Grade Long Running Agents
|
|
Trade-Off: Ease vs Control
|
|
Open-Source Protocols & Standards
|
|
MCP Donation to Linux Foundation
|
|
Challenges: Complexity, Security, Performance
|
|
Institutional Push for AI Standards
|
|
Adoption Uncertainty & Ecosystem Impact
|
|
Tiny AI Models & Edge Deployment
|
|
Aggressive Quantization Techniques
|
|
llama.cpp Framework
|
|
Vision-Language Fusion Models
|
|
Qwen 3 Series: Low-Power High Performance
|
|
Enabling AI on Raspberry Pi & Similar
|
|
Emerging Research Directions
|
|
Adaptive Agent Handoff & Orchestration
|
|
Balancing Protocol Standardization & Research Agility
|
|
Model Compression & Cross-Modal Integration
|
|
Middleware Design Patterns for Agent Loops
|
|
Secure, Performant Multi-Agent Protocols
|
|
```
|
|
- |