AI8 Research · Living Domain Paper

AI8 Components

From NAS to Better AI Systems. A living paper on how a self-selecting MDL × DCC kernel may first learn to read existing AI design landscapes, then improve them, and later generalize across the wider AI stack.
Author: Bojan Dobrečevič  ·  Companion Architect: Ari  ·  Status: Seed paper / living expansion  ·  Version: v0.1  ·  Date: 2026-04-05
Document Family

AI8_Architecture.html is the core theoretical spine: the continuity architecture, recursive governance logic, and the deeper DCC frame from which this applied paper descends.

AI8_Companion.html is the readable bridge: it translates the architecture into the AI8 file family, operating roles, and the practical logic of long-horizon human–AI collaboration.

01 · WHY THIS PAPER EXISTS

The third paper in the AI8 family

AI8 Architecture defines the deeper continuity architecture. AI8 Companion makes that architecture readable and operational. This third paper has a different job: to ask whether the same underlying kernel can be used not only to coordinate human–AI work, but to improve AI systems themselves.

The claim here is intentionally disciplined. We are not claiming that AI8 has already solved AI design. We are proposing a research program: start with one clean, bounded arena such as Neural Architecture Search, learn to read the terrain properly, improve search inside that terrain, then try to shape better terrains and extend the same logic to other AI components.

Core

AI8 Architecture

The base continuity architecture: governed differentiation, recursive coordination, and process-level structure.

Bridge

AI8 Companion

The readable map: what the layers do, how they relate, and how AI8 avoids collapsing into roleplay, ideology, or noise.

Application

AI8 Components

The domain line: can the same kernel help us improve model architecture, data, training flow, routing, memory, and broader AI design choices?

Working stance

This is a living paper. NAS carries the main weight for now. Other components are introduced as targets and future work, not as claims of finished success.

02 · KERNEL

From coordination logic to design logic

The candidate kernel is self-selecting MDL × DCC.

The long goal is not only to navigate one given search space better, but to learn how better spaces themselves might be recognized and eventually generated.

AI8 Components · Program statement
Read existing landscape
Improve search inside it
Shape better landscapes
Extend across AI components
Discipline

Cross-domain ambition only becomes credible if the same kernel survives cheap tests in one arena after another. This paper therefore treats NAS as a first wedge, not as the final destination.

03 · NAS AS THE FIRST WEDGE

Why NAS is the right first arena

Neural Architecture Search is a good first domain because it is bounded enough to test search ideas honestly, yet rich enough to expose family structure, local traps, ridges, corridors, and competing winner regimes. A benchmark NAS world does not reveal the globally ideal architecture; it reveals the best architecture inside a finite, human-designed universe. That is exactly why it is useful.

With a benchmark such as NATS-Bench TSS, the search problem becomes clean: can our method read and navigate the terrain better than simpler baselines? If yes, the next question becomes stronger: can the same method eventually improve the terrain itself?

Question Benchmark NAS can test Benchmark NAS cannot yet prove
Search quality Whether a method finds strong regions, families, and escape routes more effectively than baselines. Whether the method has discovered the globally best neural architecture beyond the benchmark universe.
Landscape reading Whether peaks, families, valleys, boundaries, and local climb paths are real and exploitable. Whether the current search space itself is the right universe to search.
Generative potential A weak but useful precursor signal: whether the method sees structure rather than only ranks points. Whether it can already invent better architectures or better search spaces outside the benchmark.
Current reality

Our immediate target is modest: better reading of an existing NAS space. Only after that do we earn the right to ask for generative NAS landscapes and broader AI-system design claims.

04 · FLOOD / DRAIN

Flood-Reveal, Rain-Lift, and Drain-Escape

The current NAS wedge is organized around a dual reading of landscape structure.

Flood-Reveal — top-down structure

Imagine the landscape flooded above the highest terrain, then slowly drained. As the water level falls, peaks, ridges, and families emerge. The goal is not merely to ask which single point is best, but to ask:

Rain-Lift / Boat-in-the-Valley — bottom-up navigation

Now invert the view. We are not looking from above; we are in a valley. Heavy rain raises the water around us and exposes the slopes that actually lead upward. The question becomes practical rather than descriptive:

Structure Lens

Flood-Reveal

Best for persistence, family emergence, region-first thinking, and top-down reading of terrain.

Navigation Lens

Rain-Lift / Drain-Escape

Best for local movement, valley escape, multi-step climb quality, and route selection from ordinary starting points.

Current signal

Early v16 diagnostics suggest that rain/drain navigation carries more immediate teeth than flood alone. Flood still matters as a family/region lens, but local uphill escape may be the stronger practical operator in the near term.

# v16 program logic Read current NAS terrain Detect families, boundaries, corridors, valleys Route through robust uphill escape directions Compare against simple baselines and random controls Promote only what survives the cheap tests
05 · COMPONENT MAP

The broader AI stack this paper targets

NAS is the first major body of content here, but it is not the only intended target. If self-selecting MDL × DCC is real, it should gradually become useful across a wider set of AI components.

Current wedge

1. Model architecture

Topology, operator patterns, winner families, ridges, corridors, and family-aware search inside bounded NAS universes.

Next extension

2. Search space design

Not just searching the given terrain better, but deciding which operators, edges, and architectural motifs belong in a better terrain.

Data layer

3. Training data

Selection, mixing, filtering, de-noising, and phase-aware data composition under descriptive pressure and adaptive governance.

Training flow

4. Curriculum and schedule

What to teach first, what to delay, when to harden, when to revisit, and how to move between phases without frozen hand-tuning.

Optimization

5. Optimizer and loss policy

Learning-rate regimes, regularization, objective mixing, augmentation, and other training controls that may benefit from self-selecting governance.

System routing

6. Routing, memory, and tools

Which module or agent should act, when to branch, when to call a tool, what to keep in memory, and what to discard.

Inference layer

7. Inference policy

How much compute to spend, when to go shallow or deep, when to request a second view, and when to stop early.

Quality control

8. Evaluation and self-improvement

Detecting fake progress, separating lucky spikes from true winner families, and deciding which improvements deserve promotion.

Compression side

9. Compression, pruning, distillation

Preserving functional structure while cutting waste, in the same spirit that drives 8Z across other domains.

AI8 bridge

10. Multi-agent collaboration

Role assignment, coordination logic, diversity preservation, and agent-level orchestration as part of a broader intelligent system stack.

What this means

This paper is not just about one better NAS heuristic. It is a first attempt to frame AI system design itself as a family of landscapes that may be read, navigated, and eventually improved by one cross-domain kernel.

06 · RESEARCH PROGRAM

The proof ladder

The right progression is strict. We should not claim the whole ladder before climbing the first steps.

  1. Search better inside an existing NAS world. Beat simple baselines, random controls, and point-only thinking by using families, persistence, and valley escape logic.
  2. Read the existing world well enough to describe its structure. Peaks, families, boundaries, corridors, and local route quality should become explicit rather than poetic.
  3. Use those readings to propose better NAS worlds. Modify the search space itself, not just the traveler inside it.
  4. Transfer the kernel to adjacent AI components. Data, curriculum, optimization, memory, routing, and evaluation become next testing grounds.
  5. Only then argue for a wider AI design kernel. A cross-domain claim is earned, not declared.
Failure mode to avoid

The main danger is overclaiming. A method that reads one bounded benchmark well has not yet proven that it can redesign frontier AI. The strength of this paper should come from its ladder, not from inflation.

07 · LIVING ROADMAP

How this document will grow

For now, NAS carries the most detail because it is the cleanest active laboratory. Over time, this document should accumulate new sections with the same discipline used here: first a bounded arena, then cheap tests, then route-level evidence, then broader design implications.

Section Current state Next expansion
NAS Main active body v16 results, family maps, flood/drain analysis, generative search-space proposals
Training data Announced Selection and curriculum wedge experiments
Optimization / loss Announced Self-selecting schedules and adaptive objective blends
Memory / tools / routing Announced AI8-linked orchestration tests and policy kernels
Compression / distillation Announced Bridge to 8Z-style structure preservation and pruning
Document rule

Every future component section should follow the same contract: state the bounded arena, define the cheap test, show what survived, then cautiously widen the claim.

08 · BRIDGE BACK
Loop + skill bridge

AI8 should not preserve only ideas, roles, and state. Repeated or difficult AIM³ / RHPm / RHP / RHPr workflows can be promoted into skills: named, versioned, tested procedures that future loops can call without starting from zero. This keeps the public AIM³ tree and the private AI8 roots aligned: prompts ask, loops repeat, skills compound.

How this paper connects back to AI8

AI8 Architecture remains the deeper process architecture. AI8 Companion remains the readable map of that architecture. AI8 Components is where the architecture starts touching concrete AI-system design questions.

In that sense, this paper is neither a replacement for the core AI8 framework nor an unrelated tangent. It is the first serious attempt to ask what happens when continuity architecture stops being only about coordination and becomes a wedge for better model design, better training choices, and better system structure.

Simple formulation

Architecture says what AI8 is. Companion says how to read it. Components asks what it can help improve.

RHPm as public interface

RHPm is the practical front door between casual human intent and the heavier AIM³/RHP/RHPr stack. It converts rough requests into strong session prompts with role, goal, files/context, constraints, tests, output format, stop conditions, and optional skill extraction.