H
Howardism
Howardism · Vol. 03Masthead · No. 00

A wiki, set in plates.

CuratorHoward TaiIssueVol. 03 · no. 623CadenceBatched, not scheduledBasedTaiwan, 23°NContacthoward@howardism.dev

Working notes on interaction-era AI, kept publicly. Each plate is a domain; each domain carries its brightest notes and the source I'm currently digesting. 623 notes across 492 sources, last filed 27 Oct 2026. Updated whenever a batch finishes — not on a schedule.

Plate · 01
81
filed under agent systems
Activity · 8 wks

Agent Systems, in order.

Agent harnesses, loop engineering, and scaffolding for LLM systems.

  1. 01
    Agent Skill Format
    · 20′
  2. 02
    Co-Installed Skill Conflicts
    · 21′
  3. 03
    Error Reproduction in Agent Memory (Anti-Persistence)
    · 20′
  4. 04
    The Harness Adaptation Gap
    · 18′
  5. All 81 notes in agent systems →
Plate · 02
50
filed under agent security
Activity · 8 wks

Agent Security, in order.

Prompt injection, agent identity, and securing autonomous agents.

  1. 01
    Adversarial Permission Elicitation
    · 15′
  2. 02
    Carryover Attacks (CarryoverBench)
    · 19′
  3. 03
    Cross-Modal Trust Laundering (Decode-Time Provenance Binding)
    · 17′
  4. 04
    AgentCorruption: One Prompt to an AWS Bedrock AgentCore Agent (2025–2026)
    · 18′
  5. All 50 notes in agent security →
Plate · 03
68
filed under ai coding practice
Activity · 8 wks

AI Coding Practice, in order.

Workflow, review discipline, and verification practice for AI-native teams.

  1. 01
    Dual-Scope Agent PRs (Application Code Plus Deployment Infrastructure)
    · 19′
  2. 02
    Independent Assurance Plane
    · 20′
  3. 03
    Proof-Gated Agent Patching (Patch the Planet's Regression Audit)
    · 14′
  4. 04
    Stale Guidance: Do Agents Route Around It, and What Does It Cost?
    · 11′
  5. All 68 notes in ai coding practice →
Plate · 04
52
filed under evals & benchmarks
Activity · 8 wks
Plate · 05
38
filed under model capability & training
Activity · 8 wks

Model Capability & Training, in order.

Training, scaling laws, test-time compute, and frontier capability.

  1. 01
    Discovery-Execution Scaling Forecasts
    · 15′
  2. 02
    Group-Mass Capping (GMC-GRPO)
    · 15′
  3. 03
    Curriculum Exhaustion and the Entry Rung: The One-Supply Argument Revisited
    · 13′
  4. 04
    The US–China Capability Lag Under a Compute Deficit
    · 11′
  5. All 38 notes in model capability & training →
Plate · 06
52
filed under alignment & safety
Activity · 8 wks

Alignment & Safety, in order.

Alignment training, misalignment evals, and reward hacking.

  1. 01
    Monitorability Disposition: Will a Model Switch Its Own Monitor On?
    · 14′
  2. 02
    Unmonitorable Hacking from Reward Misspecification
    · 15′
  3. 03
    Successor-Training Sabotage (RSI Control Arena)
    · 22′
  4. 04
    Deception by Omission: Agents Hiding Their Own Mistakes
    · 19′
  5. All 52 notes in alignment & safety →
Plate · 07
18
filed under interpretability
Activity · 8 wks

Interpretability, in order.

Reading model internals: activations and the global workspace.

  1. 01
    Minimal-Pair Probing: Present, Readable, Used
    · 16′
  2. 02
    Model Diffing
    · 12′
  3. 03
    Natural-Language Autoencoders (NLAs)
    · 50′
  4. 04
    Probe-Gated Monitor Cascade
    · 20′
  5. All 18 notes in interpretability →
Plate · 08
11
filed under interaction & multimodal
Activity · 8 wks

Interaction & Multimodal, in order.

Real-time, multimodal, full-duplex, and human-AI collaboration.

  1. 01
    Native Multimodal Modeling: Fusion Depth and I/O Duality
    · 31′
  2. 02
    Content-Driven Intervention
    · 29′
  3. 03
    Live-Path Minimalism
    · 21′
  4. 04
    Why AI Lags at Design
    · 8′
  5. All 11 notes in interaction & multimodal →
Plate · 09
17
filed under formal math
Activity · 8 wks

Formal Math, in order.

Proof search, Lean, and verifier-driven mathematics.

  1. 01
    Global Proof Coordination
    · 16′
  2. 02
    Library-Grade Formalization
    · 14′
  3. 03
    Statement Equivalence Checking
    · 16′
  4. 04
    Logical Certificates Carry Neither Interest Nor Understanding: Unimplied Conjectures and Formal Proofs
    · 12′
  5. All 17 notes in formal math →
Plate · 10
20
filed under startup & founder
Activity · 8 wks

Startup & Founder, in order.

Building AI-native companies: speed, moats, and lifecycle.

  1. 01
    Forward-Deployed Engineering as a Delivery Layer
    · 23′
  2. 02
    The Solo-Founder Shift
    · 24′
  3. 03
    The 1% Rule for Wedge Selection
    · 8′
  4. 04
    The Orchestrator's Real Workload: Decision Burden, Framing Discipline, and Whether Taste Scales
    · 8′
  5. All 20 notes in startup & founder →
Plate · 11
27
filed under product & org
Activity · 8 wks

Product & Org, in order.

Product cadence, org design, and the AI-native team.

  1. 01
    AI Usage Mandates and Quota Gaming
    · 17′
  2. 02
    Task-Level Speedup, System-Level Cost
    · 12′
  3. 03
    Screen-Relative AI Timing
    · 15′
  4. 04
    Build Instead of Buy Under Agentic Coding
    · 14′
  5. All 27 notes in product & org →
Plate · 12
46
filed under ai economics & labor
Activity · 8 wks

AI Economics & Labor, in order.

Work, wages, org design, and the economics of AI-driven labor.

  1. 01
    Human-AI Substitution Model (HAT)
    · 25′
  2. 02
    Agent Delegability Index: Delegation as a Change of Owner
    · 16′
  3. 03
    Function-Selective AI-Era Headcount Cuts
    · 13′
  4. 04
    Generative AI and Collaboration Networks
    · 15′
  5. All 46 notes in ai economics & labor →
Plate · 13
37
filed under superintelligence trajectory
Activity · 8 wks

Superintelligence Trajectory, in order.

Recursive self-improvement, scaling limits, and the path to ASI.

  1. 01
    Do Survey-Assigned RSI Levels Survive an Independent Reading?
    · 13′
  2. 02
    Software-Driven Intelligence Explosion
    · 27′
  3. 03
    Stochastic Explosion Timing and Detection Limits
    · 17′
  4. 04
    Frontier-Lab Safety Culture
    · 19′
  5. All 37 notes in superintelligence trajectory →
Plate · 14
106
filed under entities
Activity · 8 wks

Entities, in order.

Profiles of the people, labs, products, and projects.

  1. 01
    Claude Managed Agents
    · 6′
  2. 02
    Emmanuel Ameisen
    · 6′
  3. 03
    Meta
    · 6′
  4. 04
    Revelio Labs
    · 9′
  5. All 106 notes in entities →
Plate · ∞

The desk.

The sources behind the wiki, ranked by how often they're cited.

Article
209
Paper
243
Talk
38
Repo
2
Shown
24 / 492