Kaizen Craft

Nature as Algorithm

Network robustness, pollinator diversity, and learned embeddings of the waggle dance

Field Notes

Working notes on AI systems, safety research, labor, ethics, and governance policy

Literature Review of Core Quantization and Multilingual Safety Anchors

25 published papers across 17 venues (2023-2026), merging Lit Review 1.0 and 1.1.

Red Teaming Quantization growing

Jul 2026

Research Log: Redteaming Quantized Models with Multilingual Inputs

Status report: measuring Δ_HL (low- vs high-resource jailbreak ASR) under GGUF and related quantization arms.

Red Teaming Quantization growing

Jul 2026

Research Log Terms: A Plain-Language Primer

Plain-English glossary for the technical terms in Lit Review 1.0 and 1.1: quantization, GGUF, refusal directions, middle layers, Δ_HL, and related measurement terms.

AI Safety Quantization growing

Jul 2026

Frontier AI Governance Ecosystem Map

Frontier AI is governed through a chain of corporate controls, evaluators, standards bodies, regulators, national-security authorities, and international coordination mechanisms, but the weakest connection remains the step from detecting a dangerous capability to compelling action.

AI Governance BlueDot Impact growing

Aug 2026

How Do We Measure an AI Agent That Can Reason About the Test?

METR’s time-horizon work is evolving into a broader measurement program for autonomous agents: task length, economic returns, evaluation integrity, and whether AI can accelerate AI R&D.

AI Safety AI Governance seeding

Aug 2026

Critical Weight Protection for Fairness and Safety under Quantization

ACL 2026 Findings: static/dynamic PTQ degrades fairness and MultiJail safety (worse in KO/AR); Critical Weight Protection keeps top FAIRSCORE+SAFESCORE weights in FP16.

AI Safety Quantization growing

Jul 2026

See all field notes

Sketches

Experiments at different development stages

Growing Wild Gardens from The Nature of Code

Expanding on Daniel Shiffman's The Nature of Code with an interactive playground for real data.

Data Art seeding

Mar 17

CivicSpark AI

AI-enabled chat interface for Tulsa City Council. RAG over agendas, minutes, and budgets, plus topic alerts and district email drafting.

Civic Tech growing

Jul 19

Digital Brain: Local RAG + Regenerable Wiki

Local-first personal knowledge stack: Obsidian vault → Postgres/pgvector → Ollama embeddings and chat → entity graph → Karpathy-style regenerable wiki, plus a daily brief.

Data Art growing

Jul 21

See all sketches

Map

Navigation across data landscapes

Waggle Dance

The honeybee waggle dance modeled algorithmically: a location encoded as a vector, decoded by follower bees. An animal embedding. Built with Nature of Code principles.

Machine Learning Physics Simulation seeding

Seismograph Cloth

A taut verlet cloth pinned along its border, struck by a week of real earthquakes so global seismicity replays as ripples. From The Nature of Code, Chapter 6: Physics Libraries.

Physics Simulation seeding

River Tree

A real dendritic river network from OpenStreetMap around the Colorado-Green confluence, drawn by waterway class. From The Nature of Code, Chapter 8: Fractals.

Fractal Geometry Geospatial Mapping seeding

Tide Harmonics

Measured tide height at San Francisco with a two-constituent harmonic model (M2 plus S2). From The Nature of Code, Chapter 3: Oscillation.

Physics Simulation seeding

See all maps

Field Notes

Working notes on AI systems, safety research, labor, ethics, and governance policy

Literature Review of Core Quantization and Multilingual Safety Anchors

25 published papers across 17 venues (2023-2026), merging Lit Review 1.0 and 1.1.

Red Teaming Quantization growing

Jul 2026

Research Log: Redteaming Quantized Models with Multilingual Inputs

Status report: measuring Δ_HL (low- vs high-resource jailbreak ASR) under GGUF and related quantization arms.

Red Teaming Quantization growing

Jul 2026

Research Log Terms: A Plain-Language Primer

Plain-English glossary for the technical terms in Lit Review 1.0 and 1.1: quantization, GGUF, refusal directions, middle layers, Δ_HL, and related measurement terms.

AI Safety Quantization growing

Jul 2026

Frontier AI Governance Ecosystem Map

AI Governance BlueDot Impact growing

Aug 2026

How Do We Measure an AI Agent That Can Reason About the Test?

METR’s time-horizon work is evolving into a broader measurement program for autonomous agents: task length, economic returns, evaluation integrity, and whether AI can accelerate AI R&D.

AI Safety AI Governance seeding

Aug 2026

Critical Weight Protection for Fairness and Safety under Quantization

ACL 2026 Findings: static/dynamic PTQ degrades fairness and MultiJail safety (worse in KO/AR); Critical Weight Protection keeps top FAIRSCORE+SAFESCORE weights in FP16.

AI Safety Quantization growing

Jul 2026

See all field notes

Map

Navigation across data landscapes

Waggle Dance

The honeybee waggle dance modeled algorithmically: a location encoded as a vector, decoded by follower bees. An animal embedding. Built with Nature of Code principles.

Machine Learning Physics Simulation seeding

Seismograph Cloth

A taut verlet cloth pinned along its border, struck by a week of real earthquakes so global seismicity replays as ripples. From The Nature of Code, Chapter 6: Physics Libraries.

Physics Simulation seeding

River Tree

A real dendritic river network from OpenStreetMap around the Colorado-Green confluence, drawn by waterway class. From The Nature of Code, Chapter 8: Fractals.

Fractal Geometry Geospatial Mapping seeding