Tag: AI Architecture
All the articles with the tag "AI Architecture".
-
Two Classes of Agents: Codebase-Native vs Workers-Native
Agent frameworks split cleanly into two classes based on tool access, determining whether they belong in containers or on edge runtimes.
-
Context Is a Harness Artifact
LLMs are stateless. A conversation is a UX fiction. Long-horizon autonomy is a question of harness design, not model capability.
-
Dreaming and the Effect Gate
Why an active-memory substrate needs two execution modes — awake and dream — and a hard gate on effectful pipelines in dream mode.
-
Goal Generation Is Agency
Agency is the capacity to generate goals from ambient state — not the capacity to execute given goals.
-
Memory as Lazy Queries Over the World
In a stateful LLM system, memory is not a mirror of the world — it is a graph of lazy queries that resolve into the world on demand.
-
Thinking Is Substrate Self-Modification
Thinking is not something a model does. It is something a substrate does to itself, using models as workers.
-
The Harness Is a Prompt Compiler
Every LLM system is implicitly a compiler. Given accumulated history and a goal, it produces a prompt. The quality of that compiler is the entire game.
-
Prime: A Conversational Control Plane
The human talks to the business. The business talks to the code. Every conversation makes the autonomous organization smarter.
-
The Three-Layer AI Agent Architecture
How to separate agent state from LLM logic and API metering to build intelligent, financially observable agents.
-
Prime: Persistent Org-Level AI Agents on Cloudflare
A control plane architecture for autonomous multi-repo organizations — built on Cloudflare Agents SDK, Durable Objects, and GitHub as long-term memory.