# SPOTLIGHT
# ARTICLES
The Address of Authorship
How decision authority can be distributed when writing with AI, and why the tool for it looks like a text adventure
Application: Schreibwerkstatt - An ACF Writing Process for Papers, Monographs, Reportage, Essays and Books
Interactive writing workbench (German UI): a human/AI allocation mode for every writing step, an Eco-style card index, the manuscript walked as a text adventure with a consistency check, and an AI discussion partner
Application: MCAT - Human-AI Task Allocation with the Adaptive Complementarity Framework
Interactive demo of MCAT, a Human-AI task-allocation engine: run simulations, explore allocation decisions, Markov blankets, and free-energy minimization
The Strangler Strangler Flow
The Strangler Strangler Flow (hereafter: the AI Strangler Flow) reimagines legacy system displacement for the agentic AI era. Unlike traditional big-bang rewrit...
GARA: Governance-Aware Routing Architecture
Every major AI routing system optimizes the same question — which model handles this most efficiently — and none of them treat a different question as a first-c...
Beyond Capability: The Normative Boundaries of Human-AI Collaboration
Kant shows that someone must set the goals, Gödel that someone must evaluate the system from outside, and Wittgenstein that some practices dissolve entirely when the human is removed: together they establish that the human role in AI-enabled work is structural, not transitional.
A Software Architecture for Deterministic Post-Generation Validation of LLM Outputs
Logic-Guard-Layer (LGL) is a software architecture designed to close the structural reliability gap in LLM outputs without modifying the underlying model.
Application: Adaptive Multi-Agent Negotiation Framework for Decentralized Markets
The Demo of the Multi-Agent Energy Market Simulation with Lightning Network payments and Mean-Field Game Theory (see the paper)
The Art of Not Being Fooled by Your Own AI
A slightly irreverent guide to neuro-symbolic validation of Large Language Model outputs
Application: Logic Guard Layer for Deterministic Post-Generation Validation of LLM Outputs
The Demo Implementation of the Logic Guard Layer Whitepaper
Adaptive Multi-Agent Negotiation Framework for Decentralized Markets
This whitepaper presents a decentralized negotiation framework for local energy and resource markets and guides the reader from first principles to implement...
The Rule That Forgot Time
The most famous moral rule ever written overlooks one important aspect: time. This is the crack that opens when you point it at AI, the small fix that closes it, and the one seat no equation will ever fill.
# SYSTEM INFO
ARTICLES: 12
LAST ADDED: 2026-09-17