We use cookies to understand how you use this site and improve your experience.

Alexandru Mareș@allemaar
Alexandru Mareș
  1. Home
  3. Clusters
  5. Yon
Email
RSS
YounndAIYou and AI, unifiedBuilt withNollamaNollama
Cluster · active

Cluster — YON


# Cluster: YON

## Short definition

The Cluster of work covering **YON** — the constructed notation for AI cognition that separates grammar from vocabulary and treats text as a stream-first, line-independent record format.

## Long explanation

YON is the notation through which YounndAI cognitive systems read, think, and emit structured records. Unlike free-form prompts (no structure), JSON (rigid blocks that fail catastrophically on a single bad token), or YAML (block-shaped, indentation-sensitive), YON is **stream-first**: each line is a self-contained record that can be parsed, validated, or acted upon as it arrives. This matches how language models actually generate text and removes a class of failure modes that block formats inherit by design.

The other distinctive move is **grammar / vocabulary separation**. The grammar is fixed and minimal — a small structural language describing tags, fields, lists, references. Vocabularies plug in per-domain (`yon.sai`, `yon.content`, `yon.tasks`, ...). One parser, many domains. This Cluster collects every Body about YON: the spec, the parser benchmarks, the application essays, the comparisons with other agent notations.

## Why it matters

Notation shapes what you can think. Block formats forced an entire generation of LLM systems to produce-then-parse-then-fail — and to design around that failure mode (token-by-token streaming hacks, partial-JSON repair, retry loops). YON is what happens when the format *matches* the generative process rather than fighting it. The first-order benefit is correctness; the second-order benefit is composability — domain-specific cognitive vocabularies can ship on top of one parser.

This Cluster is the **language layer** of YounndAI's product line (sibling Org Alex owns). Every other concept (Elastic Automators, SAI Academy) eventually flows through YON when implemented at the cognitive-systems level.

## Best starting point

1. **The Block Problem essay:** [[2026-E0001 - The Block Problem/_metadata|E0001 — The Block Problem]] — why block formats are mismatched to LLM generation.
2. **Notation as Alignment:** [[2026-E0015 - Notation as Alignment/_metadata|E0015 — Notation as Alignment]] — the design stance behind YON.
3. **One Line One Thought:** [[2026-E0017 - One Line One Thought/_metadata|E0017 — One Line One Thought]] — the line-independence principle.
4. **Then:** browse the related Bodies below.

## Main paper / article / repo

- **Project hub:** [[yon.younndai.com|YON]] (`alm-os/Projects/YON - yon.younndai.com/`)
- **Domain:** [yon.younndai.com](https://yon.younndai.com)
- **Concept card:** [[yon|/concepts/yon]]
- **Spec / parser:** under development (Apache 2.0)

## All related Bodies

Bodies in this Cluster:

- [[2026-E0001 - The Block Problem/_metadata|E0001 — The Block Problem]] (2026-04-01) — foundational mismatch motivating YON
- [[2026-E0014 - The Strong Form/_metadata|E0014 — The Strong Form]] — argument for explicit structure
- [[2026-E0015 - Notation as Alignment/_metadata|E0015 — Notation as Alignment]] — why notation choice shapes cognition
- [[2026-E0016 - The Borges Warning/_metadata|E0016 — The Borges Warning]] — adjacent: language and meaning
- [[2026-E0017 - One Line One Thought/_metadata|E0017 — One Line One Thought]] — line-independence principle
- [[2026-E0025 - Why Constructed Languages Always Failed - Until Now/_metadata|E0025 — Why Constructed Languages Always Failed — Until Now]] — situates YON in the longer history of constructed languages
- (More Bodies as the Arc continues.)

## Videos / diagrams / infographics

- Per-episode video shorts; permalinks captured in each episode's `_metadata.md`.
- Future: notation-comparison diagrams; YON parse-tree visualizations.

## External references

- IndieWeb POSSE — adjacent canonicalization stance for human content.
- Wirth-style grammars and EBNF — the lineage YON's grammar layer draws on.
- Domain-specific languages literature (Fowler, Spinellis) — the vocabulary-pluggability analogue.

## Related topics

- [[ai-cognition|Cluster: AI Cognition]] — what YON enables at the cognition layer
- [[elastic-automators|Cluster: Elastic Automators]] — Elastic Automators expressed in YON
- Constructed-language history (Cluster TBD)

## FAQs

**Q. Isn't YON just YAML with a different name?**
A. No. YAML is block-structured, indentation-sensitive, and fails on partial parses. YON is stream-first, line-independent, with a grammar-vocabulary split absent from YAML.

**Q. Does YON replace JSON?**
A. For LLM-driven cognition: yes, by design. For static configuration files where the whole document is produced atomically: JSON is fine; YON's design pays its costs only where streaming matters.

**Q. What's the relationship to JSONL?**
A. Closer than to JSON or YAML. YON shares JSONL's line-as-unit move but adds a grammar layer (records have typed tags, not anonymous JSON objects) and the grammar-vocabulary split.

## Latest updates

- *(in development)* — YON v1.0 spec.
- *(in development)* — `@younndai/yon-parser`.
- Episodes E0014–E0017, E0025 published.

Member Bodies (13)

  • Why Constructed Languages Always Failed — Until Now
    2026-04-24

    Every language you speak was an accident. No one designed English.

  • One Line, One Thought
    2026-04-19

    Every line should be a complete thought. One rule. It changed computing fifty years ago.

  • The Borges Warning
    2026-04-18

    The language you give an AI decides what it can think. What it can't think... doesn't exist.

  • The Strong Form
    2026-04-13

    Imagine you can hear shapes. A bat can. Echolocation. Every surface in a room has a sound.

  • Probably Purple
    2026-04-12

    Someone's writing a brief. Typing it out. Not thinking about formats or systems.

  • Storage vs Orchestration
    2026-04-11

    A recipe card tells you what's in the dish. It doesn't cook dinner.

  • Walls and Doors
2026-04-09

The problem isn't that industries speak different languages.

  • The Block Problem
    2026-04-03

    Imagine writing an entire essay, but you can never re-read what you already wrote.

  • Teaching the Machine: YON as a Curriculum for AI
    2026-02-24

    We are not building calculators anymore. We are raising minds. The syntax we choose is the lesson we teach. If we choose chaos, we teach confusion. If we choose structure, we teach clarity.

  • Agent Economy: Trust and Transactions
    2026-02-19

    Text is cheap. Action is expensive. We have spent a decade teaching machines to speak. The next economy is about what AI does. The barrier is not intelligence. It is trust.

  • 226 Milliseconds
    2026-02-17

    I stared at the benchmark results. The parse speed was ten times slower than JSON. I should have been horrified. Instead I was smiling. I was looking at the wrong number.

  • Anatomy of a Fail-Closed System
    2026-02-10

    I built the Runner to say no. Not because I distrust the machine. Because I respect the weight of action.

  • JSON is Dead: Why We Need to Stop Speaking 1999
    2026-01-05

    It is 2 AM on a Tuesday. The system fails. It fails because of a missing comma. We are trying to run advanced intelligence on a file format designed for a blog comment section.