February 20, 2025

Author - John Deacon

John is a researcher and practitioner committed to building aligned, authentic digital representations. Drawing from experience in digital design, systems thinking, and strategic development, John brings a unique ability to bridge technical precision with creative vision, solving complex challenges in situational dynamics with aims set at performance outcomes.

Understanding embeddings in vector spaces and their impact on accuracy and nuance in prompt engineering

107 views

5 min read

Vector spaces, their relationship to embeddings in LLMs, the meaning of dimensions, the algorithms that generate embeddings, and their role in accuracy and nuance in prompt engineering. 1. What Is a Vector Space in the Context of AI...

Managing Context in Real-Time AI Retrieval with Semantic Embeddings

105 views

4 min read

When building a large contextual database for real-time AI retrieval, using semantic embeddings can greatly simplify managing nuanced context. However, directly reading and writing to embeddings comes...

Mapping the CAM Objective Function with Real-Time Observer Focus

102 views

3 min read

Mission Layer (Core Purpose and Goal Alignment) Core Purpose: The CAM Objective Function emphasizes the alignment of real-time observation with the deeper layers of human cognition and perception. By engaging methodologies that...