H
Howardism
Howardism · Vol. 03Plate I · No. 01

Writing, in order.

Pieces373Sections4Oldest10 Apr 2026Newest12 Aug 2026

Every article in the wiki, grouped by kind: Concept notes, Entity profiles, and Essay pieces. Hover any title for a preview; click to enter.

Filed under
allConceptEntityEssayIndex
Domain
agent systemsagent securityai coding practiceevals & benchmarksmodel capability & trainingalignment & safetyinterpretabilityinteraction & multimodalformal mathstartup & founderproduct & orgai economics & laborsuperintelligence trajectoryentitiessyntheses
Plate · 01 of 04
247
concept notes

Concept, in order.

Product & Org17
  1. 01
    Prototype Fidelity After Cheap Polish
    Design ProcessPrototypingResearch Agenda+1
    · 10′
  2. 02
    Standardize the Infrastructure, Not the Tools
    AI Native OrganizationEngineering LeadershipPlatform+1
    · 6′
  3. 03
    Community Smells Under AI Adoption
    Team DynamicsOrganizationCollaboration+2
    · 12′
  4. 04
    Excellence as an Operating System
    CultureTalent DensityOrg Design
    · 6′
  5. 05
    Systems Thinking Over Specialization
    Team DesignHiringPlatform Engineering+1
    · 11′
All 17 →
Plate · 02 of 04
72
entity notes

Entity, in order.

  1. E01
    LlamaIndex
    RetrievalDocument ParsingAgent Engineering+1
    · 5′
  2. E02
    Jeff Dean
    EntityPersonGoogle+2
    · 8′
  3. E03
    Elon Musk
    EntityPersonFounder+1
    · 9′
  4. E04
    Erik Brynjolfsson
    EntityPersonEconomist+2
    · 8′
  5. E05
    GPT-Live
    Type/entityLLM Model
    · 4′
  6. E06
    Bun
    EntitySoftwareAnthropic+1
    · 5′
  7. E07
    Cline
    EntityAgent RuntimeCoding Agent+2
    · 6′
  8. E08
    Cursor
    EntityOrganizationAI Coding+1
    · 6′
  9. E09
    Jarred Sumner
    EntityPersonAnthropic+1
    · 4′
  10. E10
    OpenHands
    EntityToolAgent Harness+1
    · 4′
All 72 entity notes →
Plate · 03 of 04
39
essay pieces

Essay, in order.

  1. S01
    Authority and Audit Survive Abundance
    DerivedHarnessAuthorization+2
    · 12′
  2. S02
    What Makes a Self-Improvement Artifact Transfer?
    DerivedAgent EngineeringHarness+2
    · 10′
  3. S03
    App Server vs MCP, and the Claude-Side Equivalent: Three Boundaries for Driving Agents
    DerivedProtocolMCP+2
    · 7′
  4. S04
    Bind, Don't Forbid; Prevent, Don't Detect: The Action-Open and Poisoned-Memory Residuals
    DerivedSecurityPrompt Injection+2
    · 6′
  5. S05
    Classifier Gates vs OS Sandboxing: The Defense-in-Depth Story for Auto Mode and Cowork
    DerivedSecurityClaude Code+2
    · 7′
  6. S06
    When Knowledge Layers Disagree: Context Files vs Memory, and Conflicting Sources at Compile Time
    DerivedKnowledge ManagementAgent Engineering+2
    · 7′
  7. S07
    The HTML Artifact Lifecycle: Where Plan History Lives, and When Disposable Becomes Durable
    DerivedHuman AI CollaborationAgent Engineering+2
    · 6′
  8. S08
    Is Human Review of AI-Authored Code Still a Real Control, or Already Rubber-Stamping?
    DerivedAI Coding WorkflowCode Review+2
    · 14′
All 39 essay pieces →
Plate · 04 of 04
15
index notes

Index, in order.

  1. I01
    Open Questions Backlog
    · 84′
  2. I02
    Interpretability
    · 4′
  3. I03
    Agent Security
    · 8′
  4. I04
    Agent Systems & Harness Engineering
    · 16′
  5. I05
    AI Coding Practice
    · 12′
  6. I06
    AI Economics & Labor
    · 10′
  7. I07
    Alignment & Safety
    · 8′
  8. I08
    Evals & Benchmarks
    · 6′
All 15 index notes →
Plate · #
386
subjects

By subject, most cited.

Agent EngineeringEntityAnthropicLLM ArchitectureEmpiricalAlignmentGovernanceAI Coding WorkflowDerivedEvaluationSecurityPersonHuman AI CollaborationWorkforceHarnessSafetyMeasurementBenchmarksStartupInterpretabilityContext ManagementEconomicsLLM ModelOpenaiProduct ManagementAI Native OrgCapability EvaluationFounderMultimodalAgent OrchestrationEngineering MetricsLLM As A JudgeOpen WeightsPrompt InjectionCapability TrajectoryClaudeEvaluation MethodologyFailure ModesOrgRecursive Self ImprovementTest Time ComputeAI CodingCode QualityCostGoogleKnowledge ManagementTheoryAI NativeAI RdCode ReviewGovernance WorkforceMonitoringOrg DesignPlanningPost TrainingThreatsZero TrustAccountabilityAI For MathematicsClaude CodeForecastingMulti AgentOpen SourcePromptingTrainingAdaptive EvaluationAgentsAI LabAI PolicyAuthorizationAutomationEvalsHonestyMCPMoatsOptimizationOrganizationPractitioner OpinionProduct OrgProduct StrategyReference MonitorReinforcement LearningReliabilityResearcherRetrievalReward HackingRole EvolutionTeam DesignToolTraining GamingTrust BoundaryVerificationWorkflow DesignAgent HarnessAgent RuntimeAI AdoptionArchitectureAsiChain Of ThoughtCognitive LoadCompetitive StrategyCybersecurityDeep ResearchDesignDeveloper WorkflowFormal MethodsGo To MarketIdentityIncident ResponseIntegrationLeast PrivilegeLLM EvaluationMacroMemoryModel WelfareOrchestrationPrdProductProduct ProcessPrototypingScalingSoftware EconomicsSuperintelligenceTasteTechnical DebtTechnology DiffusionType/entityAccess ControlAgent DeploymentAgentic RlAI CoworkingAI EvaluationAuthenticationBelief ModificationCalibrationCareerCatastrophic RiskCodexCodingCognitionCompanyCoordinationCoworkData ContaminationDeceptionDeepmindDefense In DepthDefensibilityDeploymentDocument ParsingDual UseEducationExpertiseFrontier LabsGrowth DynamicsHRInferenceInference ScalingInterfaceInterpretability ResearcherIntrospectionLaborLimitsLLM CharacterMental ModelMethodologyMetrMisalignmentModel ImprovementModel RoutingModel SpecOversightPlatformPrivacyProcessProduct ArchitectureProduct TasteProductivityProtocolProvenanceQualityQuantizationResearch MethodologyResearch ProgramRLHFRspSelf ReportSkillsSoftware DesignStandardsStartupsSupply ChainSynthesisSynthetic DataSystemTask SpecificationThreat LandscapeUniversal AIVerifiabilityVulnerability ResearchY Combinator
+185 one-off subjects
AbstractionAgencyAgent FrameworksAgent SafetyAIAI EconomyAI EducationAI EthicsAI Native OrganizationAI SafetyAI ToolsAixiAlignment ResearcherArms ControlArtifactsAsset PricingAuditingAuthorshipAutonomyBenchmarkBusiness StrategyCapabilityCausal TheoryCharacterClassifiersCLI AgentCode AuthorshipCoding AgentCoding AgentsCollaborationCompetitionCompetitive LandscapeComplexity TheoryComputeComputer UseConcept DiscoveryConfirmation BiasConfused DeputyConsciousnessConstruct ValidityContainmentContradictionControl PlaneCopyrightCreativityCredentialsCultureData FlywheelData PipelineData ProvenanceData WallDebate MapDebuggingDecision MakingDefenseDefinition Of DoneDelegationDelimiter InjectionDemocratizationDesign PrincipleDesign ProcessDesign SystemsDeveloper EndpointDigital IntelligenceDiscoveryDocumentEconomic ModelingEconomistEfficiencyEngineering IntelligenceEngineering LeadershipEpistemicsEquityExploit DevelopmentFailure ModeFaithfulnessFormal VerificationFoundersFrontier LabGame TheoryGeneralistsGeneralizationGoodhartGoogle DeepmindGrounded TheoryGroup AgencyGrpoGtmHallucinationHciHiringHistoryHouseholdHuman CapitalIncidentsInformation Flow ControlInput ValidationIntelligence MeasureInteraction MultimodalInternal ToolingInverse ScalingIsolationJailbreakKnowledge SharingLatent CapabilityLearning GuideLeverageLifecycleLLM CapabilitiesLoopsLow RankMatrix CompletionMeasurement ValidityMediaMemorizationMessaging PlatformMethodMicrosoftMidtrainingModelModel CapabilityModel PoisoningMultilingualMvpNational SecurityNetflixNous ResearchOff HostOpen Ended DiscoveryOpinionsOwaspPermissionsPersonaPersonal Knowledge BasePlatform EngineeringPm SkillPricingPrincipleProbesProductionPrompt EngineeringPropagationQualitative AnalysisRAGReasoningRed TeamingRepresentationsResearchResearch AgendaRevenue Per EmployeeReward DesignRole ShiftRubricsSafeguardsSandboxingScale StageScaling LawsScientific DiscoverySearchSFTSingularitySkills DevelopmentSoarSoftwareSoftware ArchitectureSoftware DevelopmentSoftware ParadigmStanfordSubagentsSystemsTalent DensityTeam DynamicsTestingThreat ModelTool PoisoningTool UseTraining EfficiencyUncertaintyUnit EconomicsUnlearningValidationValue ModelVendorVersioningWorkflow