context-engineering
Models should not rediscover the grammar.
A grammar that never changes belongs in the harness, not in the model's probability budget.
Shep Bryan
field-note
Google Veo 2: Video Prompting Guide
Google's Veo 2 is an AI video generator capable of transforming still images into visually compelling 8-second clips. Google's innovation is particularly impressive due to its grasp of cinematic language. Using terms like "timelapse," "aerial shot," or "side-scrolling dolly&
field-note
Shep’s Neo4j Cypher Query Cheatsheet
📝 What is Neo4j and Cypher? Neo4j is a graph database that stores data in nodes and relationships rather than tables. This structure is perfect for highly connected data where relationships are as important as the data itself. If you've ever diagrammed an idea with circles and arrows on
field-note
Shep’s Github CLI Cheatsheet
Introduction Everyone is becoming an AI engineer these days – some of us are just doing it more quickly than others. If you've arrived in the AI space without a traditional engineering background, then using GitHub is likely a new skill for you. GitHub is essential for development projects,
field-note
The Shape of Thought: Knowledgespace
Preamble: In this post I introduce a theory for the geometry of knowledge in static embedding spaces, along with some early thoughts on AI as a collaborative epistemic navigator. This is partially inspired by Peter Gärdenfors’ The Geometry of Meaning. In the AI field I would label myself a self-
Innovation Is Subjective
One thing you learn in a job focused on innovation is that innovation is highly subjective. If you're serving a client with a specific idea of how they want to innovate and you desire their money, sometimes you just accept their understanding of innovation as the operating definition
field-note
Making Claude 3.7 Sonnet Think in Unicode & Reason Semi-Symbolically
I'm a bit of a madman for this. Claude 3.7 Sonnet is out now with an extended thinking mode, but it still only thinks in plain text. So... I added a symbolic layer to the reasoning process with a Unicode Thinking Protocol that buckets different styles of
field-note
What 200+ Hours of AI Development Taught Me About Building with Lovable.dev
We’re all software engineers now. Type it and AI will make it for you. Since mid-2024, I’ve spent 200+ hours working with Lovable.dev, building over 60 different projects along the way. If you’re unfamiliar with this tool, Lovable is an AI software engineer that codes
field-note
Batshittification: A Lens for Bold Ideas in a Sterile World
Why "reasonable" ideas won't save us—and what to do instead. The Problem: A Sterile, Over-Optimized World Look around. Everything is starting to feel the same. We’ve optimized our products, services, and systems into a dull sameness: * Products have been squeezed dry by over-
AI
The Expertise Engine: A Framework for Augmenting Knowledge Work
Introduction Something unprecedented is happening in knowledge work. The gap between average and exceptional performance used to top out around 10x. A great programmer might crank out ten times the code of their peers. A stellar consultant might handle ten times the normal client load. But that ceiling has been
field-note
Minimum Viable Ontology (MVO) Framework
Introduction In the face of AI disruption, the startup ecosystem is accelerating to a pace we’ve never seen before. And while AI language models have unlocked impressive new ways of working, there is a new field of play emerging that will be a defining factor of competitive advantage going
field-note
Knowledge Diffusion: A New Paradigm in LLM Information Extraction
Introduction In my recent explorations of Large Language Models (LLMs), I've stumbled upon what I believe to be a significant breakthrough in how we can extract and organize information from these powerful AI systems. This article introduces the concept of knowledge diffusion, a novel framework for AI-driven