Essays
Longer pieces on where AROS is going and why.
Wang Ziyi26 min readEnglish
Asking a compiled literature
A search engine returns papers. A compiled literature can answer a question with a whole family of analyses, find what nothing explains, and keep score.
Wang Ziyi30 min readEnglish
From auditing one paper to compiling a field: why living meta-science needs packing that keeps the conditions
Summarising papers by their abstracts packs their errors into the consensus. Auditing each study into a capsule lets software cross-check and update a field.
Wang Ziyi43 min readEnglish
Why five axioms from mathematical logic govern how we audit science
Many theorems rest on five levels of proof. The same five levels guide how our skills audit manuscripts and how an AI agent checks its own actions.
Wang Ziyi14 min readEnglish
Beyond the chat window
Scientists work with AI by conversation, which loses the thread; private systems keep it but wall it in. We are building a shared research-experience system.
Wang Ziyi20 min readEnglish
Why we turn the rules for auditing research into AI agent workflows
Science's checks catch errors, but years late. Agent workflows that audit research run them while the work is done, helping people and AI do research together.
Wang Ziyi10 min readEnglish
Why we build workflows from verified nodes
An agent should compose tested methods, not rewrite them. What verified nodes save, what they keep the model from getting wrong, and where you decide AI works.
Nexitia1 min readEnglish
Why we write in public
What these essays and the dev log are for, and the rules we hold them to.