Weekly AI Dispatch
Week of May 11, 2026
Join the Circle
Three points. No fluff. Every Monday.
Section Two
The Stack
Personal, opinionated tool curation — not a directory. Tools used as a thinker + creator, not as a developer. Updated quarterly.
Why this works: Authenticity is the product. Curation makes the decision for the reader. This is not a comprehensive directory — it's what actually gets used in practice.
Section Three
AI Literacy & Learning Path
Take the quiz to assess your level, or dive directly into the learning path. Always accessible — no gatekeeping.
Orientation: What You're Actually Walking Into
Understand the intellectual history and conceptual structure of AI before writing a single line of code. Most people skip this. That is why most people remain confused about the difference between AI, ML, GenAI, and agents for years.
Mathematical Spine
Build the mathematical intuition required to understand why neural networks work — not just that they do. You do not need a PhD in mathematics. You need honest fluency in four areas.
Classical Machine Learning
Understand the learning paradigm — how a system extracts patterns from data — before entering the complexity of deep learning. These methods are still used in production daily. Understanding them makes you better at understanding deep learning.
Deep Learning
Understand how multi-layer neural networks learn representations from data — and why they work at all. This phase covers the fundamental building blocks of everything that comes after.
The Transformer Revolution
Understand the architecture that powers every major AI system of the modern era. The transformer is not a trend. It is the current fundamental unit of intelligence in AI systems.
Foundation Models and the LLM World
Understand how LLMs are built, trained, aligned, and deployed. Move from understanding the architecture to understanding the ecosystem — prompt engineering, fine-tuning, multimodality, and the full stack of modern AI applications.
Agentic AI
Understand AI agents — systems that can perceive, reason, plan, and act in the world across multiple steps. This is the current frontier of applied AI and the area evolving fastest.
The Hard Problems
Engage seriously with the unsolved problems in AI — the challenges that define the research frontier and that separate people who use AI from people who advance it.
Building: From User to Creator
Build production-quality AI systems. Move from understanding models to shipping systems. This is where technical fluency meets product thinking.
The Frontier
Develop the ability to track, understand, and eventually contribute to the research frontier. Not every practitioner becomes a researcher. But understanding what is happening at the frontier changes how you build and what you see as possible.
Recommended Learning Sequence
12 months of serious, consistent engagement. Every phase earns the next one.
This roadmap will evolve. The field does not stay still. The measure of whether you have internalized it is not whether you reach Phase 12 — it is whether, by Phase 9, you are starting to generate your own questions about what comes next.