Governance latency and contextual debt in AWS Context Ontology Accelerator…
To what extent does the human-in-the-loop governance requirement in the Amazon Web Services (AWS) Context Ontology Accelerator (COA) workflow exacerbate the stability-plasticity dilemma for agents con…
Decision governance for decentralized execution
How do large, established organizations deliberately design, implement, and continuously recalibrate the interdependencies among (1) decision governance systems (allocation of strategic versus operati…
Secure Runtime Evolution for AI Coding Agents
What is the logical progression in AI (Artificial Intelligence) coding-agent runtime design from local process/Operating System (OS) sandboxes, through shared Continuous Integration (CI)/cloud develop…
Hybrid memory integration
How can Artificial Intelligence (AI) agents effectively synchronize structured semantic memory, meaning ontologies and knowledge graphs, with latent knowledge encoded in Large Language Model (LLM) wei…
Episodic-to-semantic memory consolidation in AI agents
What techniques enable AI agents to reliably generalize from specific episodic experiences (interaction logs, task traces, observed events) to durable semantic memory entries (ontological facts, proce…
Autonomous knowledge curation and truth maintenance for agentic ontologies
What mechanisms exist, or are under active research, to enable Artificial Intelligence (AI) agents to autonomously curate which extracted knowledge is worth retaining in a long-term ontology, detect a…
Privacy-preserving long-term memory for Artificial Intelligence agents
How can Artificial Intelligence (AI) agents preserve the utility of long-term memory for personalisation and historical context while enforcing privacy, security, and data-sovereignty controls strong…
Symbolic-connectionist synchronisation in hybrid agent memory
How can hybrid agent-memory architectures keep structured symbolic knowledge bases synchronised with unstructured Large Language Model (LLM) and retrieval-layer memory so that updates remain consisten…
Autonomous forgetting and information curation for long-term agent memory
How can Artificial Intelligence (AI) agents implement autonomous forgetting mechanisms and information-curation policies that preserve long-term memory utility while preventing retrieval quality, late…
Evaluation frameworks for agentic memory quality, relevance, and retrieval…
What benchmark suite and metric design best measures the quality, relevance, retrieval accuracy, freshness, and governance correctness of agentic memory systems across heterogeneous tasks?
Episodic-to-semantic memory consolidation architectures for agents
What architectures most effectively consolidate raw episodic traces into reusable semantic knowledge for Artificial Intelligence (AI) agents, and which triggers, review loops, and intermediate represe…
AWS AgentCore and AWS-native Knowledge Context Layer
What Amazon Web Services (AWS) AgentCore capabilities and AWS-native services are required to design and operate a Knowledge Context Layer (KCL) that continuously acquires, curates, evolves, and serve…
How should the balance between standardized and customized internal tooling…synthesis
How should the balance between standardized and customized internal tooling shift across industries, organisation sizes, maturity levels, and Artificial Intelligence (AI) agent adoption patterns, and…
AI productivity, quality, and governance open questions
What empirical evidence can distinguish sustainable Artificial Intelligence (AI)-enabled software delivery gains from short-lived throughput effects and hidden quality or governance costs in productio…
AI-first software ecology in large engineering organisations (2025-2030)
What operating model, architecture strategy, and governance practices best improve developer productivity with Artificial Intelligence (AI) assistance in large software organisations, while preserving…
Joint Embedding Predictive Architecture (JEPA) shift
Is the shift from text-token prediction to Joint Embedding Predictive Architecture (JEPA)-style video outcome prediction the same class of problem as the shift from video prediction to physically grou…
Ontology Completeness as a World Model for Large Language Model (LLM) Prediction
To what extent can a sufficiently complete ontology function as a practical world model (in the sense described by Yann LeCun) for Large Language Models (LLMs) making predictive inferences, and which…
What capabilities, sub-capabilities, architectural patterns, and maturity…
What are the key capabilities, sub-capabilities, architectural patterns, and maturity dimensions for tool-using, semi-autonomous Semantic Knowledge Management (SKM) systems that integrate automated ha…
What is the Dynamic Resource Discovery architecture pattern in multi-agent…
What is the Dynamic Resource Discovery (DRD) architecture pattern in multi-agent systems, how does it relate to context engineering, meaning the design of what information enters an agent's working co…
What is the most practical enterprise design for a five-pillar knowledge…synthesis
What capability architecture, control model, and operating system of work best implement a five-pillar agentic, meaning tool-using and semi-autonomous, Knowledge Management (KM) model for Artificial I…
How should financial Retrieval-Augmented Generation (RAG) systems filter…
What pre-retrieval architecture and governance controls most reliably remove low-information content, meaning boilerplate, repeated passages, wrapper text, and other low-signal document fragments, and…
At what threshold does Human-in-the-Loop (HITL) oversight in bank compliance…
What measurable workload, alert-volume, and staffing thresholds indicate that Human-in-the-Loop (HITL) compliance review is no longer a meaningful challenge function, meaning reviewers mostly accept a…
How should banks detect and mitigate user-belief mirroring and sycophantic…
How do standard prompt-engineering patterns used in banking credit and compliance workflows trigger sycophancy, meaning model behaviour that agrees with user-stated beliefs over better-supported answe…
How should banks stop fluent but weakly evidenced Artificial Intelligence…
How does polished, authoritative generative Artificial Intelligence (AI) prose affect reviewer behaviour in anti-money laundering (AML) and Know Your Customer (KYC) workflows, and which interface and…
How should banks govern department-level agent sprawl and bottleneck shifts…
How does uncoordinated growth of department-level software agents change systemic risk and reporting integrity in banks, and which governance architecture can maintain consistency across agents, trace…
Flexibility vs. Predictability
In a production pipeline with uncontrolled inputs, how does the trade-off between the flexibility of an agentic system and the predictability of a deterministic execution model affect the auditability…
Stochastic LLM Agent vs. Deterministic Coded System
How do the failure modes of a stochastic multi-step Large Language Model (LLM) agent, meaning a tool-using system whose action path can vary across runs, differ fundamentally from the failure modes of…
Formal Generalisation Bounds for Tool-Using LLM Systems When Tools Return…
What formal bounds can be stated for generalisation outside the training distribution in tool-using Large Language Model systems when their tools return non-deterministic outputs under unconstrained p…
Adversarial Input Propagation Through Multi-Step Tool-Using LLM Systems
How do adversarial inputs or unexpected environmental shifts propagate error through a multi-step tool-using Large Language Model (LLM) system's verification and strategy-selection phases when the und…
Agentic Tool-Feedback Loops and Explanatory Reach
When a Large Language Model (LLM) is wrapped in an agentic loop, meaning a repeated perception, strategy-selection, tool-action, and verification cycle, does the outer loop introduce true explanatory…
What Are We Losing and Gaining by Inserting Autonomous Tool-Using Artificial…synthesis
What are we concretely losing and gaining, across the dimensions of capability, reliability, auditability, explainability, and organisational risk, by inserting autonomous tool-using Large Language Mo…
Are Multi-Step Large Language Model-Based Systems Inherently Less Explainable…synthesis
Are multi-step Large Language Model (LLM)-based systems inherently less explainable than equivalently scoped deterministic software systems, or does production-scale distributed-system complexity make…
Visibility and exit outcomes
How often does vendor-supplied temporary operational automation produce materially worse visibility and exit outcomes than internally governed temporary operational automation?
Longitudinal persistence rates after gap closure for low-code applications,…
What longitudinal evidence exists on persistence rates after the original gap is closed for low-code applications, bots, and agents in live enterprise estates?
LLM-First Policy Clarification and Institutional Knowledge Atrophy
How does shifting from peer policy clarification to Large Language Model (LLM)-first interaction affect institutional memory transfer, mentoring, and long-term policy expertise?
Policy Quality Degradation and Cross-Institution Blind Spots When New Policy…
What policy-quality degradation and systemic blind-spot risks emerge when organisations draft new policy versions from Large Language Model (LLM) interpretations of previous policy versions?
LLM Training Prior Contamination in Compliance Interpretation
What failure modes emerge when Large Language Models (LLMs) combine generic public legal knowledge with proprietary organisational policy in compliance interpretation tasks?
Cognitive Closure Under Ambiguity and Confirmation Bias
How do pressures to reach a quick, definite answer under ambiguity and iterative prompt refinement influence acceptance of flawed Large Language Model (LLM) policy interpretations?
De Facto Policy Drift From Repeated Unverified LLM Interpretations
How quickly do repeated unverified Large Language Model (LLM) interpretations create de facto policy norms that diverge from executive intent and board-level risk appetite?
Adversarial prompting risks in policy assistants
How vulnerable are corporate compliance Large Language Models (LLMs) to adversarial prompting that reframes restrictive policy as permissive guidance, and which controls detect or contain deliberate m…
AI-Assisted Policy Interpretation and Accountability Displacement
How does integration of Large Language Models (LLMs) into policy-ambiguity resolution change liability allocation, escalation behaviour, and an organisation's ability to justify the resulting decision…
Policy enforcement and formal verification as Energy-Based Model (EBM)…
How can discrete policy engines and formal verifiers be translated into continuous or structured optimization signals that guide Energy-Based Model (EBM) search while preserving the original natural-l…
Layered reasoning stack interfaces
What state abstraction boundaries and interface protocols are most effective for mapping Large Language Model (LLM) candidate outputs into Energy-Based Model (EBM) evaluation state spaces while preser…
Kona and Aleph at their core, with Lean and unifying concepts
What are Kona and Aleph at their core, what do they each do in practice, how does Lean (the theorem prover) relate to them, and which unifying concepts explain where they overlap and differ?
ServiceNow Artificial Intelligence (AI) Control Tower
What is the complete set of features, functions, and capabilities offered by ServiceNow AI Control Tower, and how do those capabilities address enterprise Artificial Intelligence (AI) governance, obse…
Microsoft Copilot Studio
What is the complete set of features, functions, and capabilities offered by Microsoft Copilot Studio, and how do those capabilities support enterprise-grade Artificial Intelligence (AI) agent develop…
Microsoft Foundry (formerly Azure Artificial Intelligence (AI) Foundry)
What is the complete set of features, functions, and capabilities offered by Microsoft Foundry, and how do those capabilities support the full Artificial Intelligence (AI) development lifecycle, from…
Amazon Web Services (AWS) Bedrock platform capabilities
What is the complete set of features, functions, and capabilities offered by Amazon Web Services (AWS) Bedrock, including its model access, agent building, knowledge bases, guardrails, evaluation, and…
Amazon Bedrock AgentCore and related suite
What is the complete set of features, functions, and capabilities offered by Amazon Bedrock AgentCore and its related suite, including AgentCore Gateway, AgentCore Memory, AgentCore Identity, and the…
Variance Control Comparison Across Delivery Modes
What is the empirical failure-rate distribution of Artificial Intelligence (AI)-assisted code that has passed a standard software delivery pipeline compared with AI-agent-executed business processes a…
Information Technology (IT) throughput capacity as a constraint on unmet…
What is the empirical relationship between Information Technology (IT) throughput capacity and the rate at which unmet operational capability needs accumulate across comparable organisations, and what…
Automated decommission of temporary bridge Artificial Intelligence (AI) agents
What technical and organisational mechanisms most reliably cause temporary bridge Artificial Intelligence (AI) agents to be decommissioned when the corresponding software capability is delivered, and…
External Dependency Surface Taxonomy for Production LLM Agents
What is the complete taxonomy of external dependencies for a production Large Language Model (LLM)-based agent, how does each dependency class fail, what is the blast radius of each failure class, and…
Temporary Automation Demand Persistence and Core Capability Investment…
What evidence exists that temporary automation workarounds displace investment in core software delivery, and what is the observed persistence rate of those workarounds after the underlying systems ca…
Agent Operational Cost vs Gap Closure Cost
What is the fully loaded operational cost of a production Artificial Intelligence (AI) agent used as a workaround for a missing system capability, relative to the cost of closing the underlying system…
Declaration of the Independence of Cyberspace
What are the historical origins and core claims of John Perry Barlow's *Declaration of the Independence of Cyberspace*, how have those claims influenced modern research and technology governance, and…
PromptQL definition, research foundations, and related technologies
What is PromptQL, what active research areas are most closely related to it, what prior research foundations PromptQL appears to build on, and which adjacent technologies should be considered when eva…
Endsley Model of Situational Awareness deep dive
What is the Endsley Model of situational awareness, meaning the perception of relevant elements, comprehension of their meaning, and projection of their near-future status, how are its three levels de…
What is Anthropic's '4D' framework for Artificial Intelligence (AI) fluency,…
What is Anthropic's "4D" framework for Artificial Intelligence (AI) fluency, what do each of the four Ds, Delegation, Description, Discernment, and Diligence, mean in practice, and how does this frame…
Architectural patterns for reliable organizational process identification,…
What integrated architectural configuration of retrieval, reconciliation, constraint enforcement, memory, validation, escalation, and governance mechanisms most reliably enables visual workflow toolin…
Agent-to-Agent (A2A)-to-tool-calling unification
To what extent does unifying specialised Agent-to-Agent (A2A) protocols into a standardised tool-calling interface affect orchestration overhead and reasoning accuracy in hierarchical multi-agent syst…
When Retrieval-Augmented Generation source documents change after agent build…
When the source documents indexed in a Retrieval-Augmented Generation (RAG) pipeline change after an agent has been built and tested, what failure modes and behavioral regressions can result in produc…
Web ontologies in production Knowledge Graphs for multi-step Artificial…
How should web ontologies, Resource Description Framework (RDF), Web Ontology Language (OWL), RDF Schema (RDFS), Simple Knowledge Organization System (SKOS), and Schema.org, be selected, composed, and…
Open Digital Rights Language (ODRL) policies in Knowledge Graphs for…
How can the World Wide Web Consortium (W3C) Open Digital Rights Language (ODRL) be used to encode access control, usage policies, and governance constraints within or alongside a Knowledge Graph (KG)…
Knowledge Graph lifecycle management for multi-step software agents
What are the best practices for maintaining and evolving a Knowledge Graph (KG), a structured graph of entities and relationships, that serves multi-step software agents, covering schema versioning, e…
Knowledge Graph as a data product
What does it mean to treat a Knowledge Graph as a data product in a data mesh architecture, and how should data product principles, including domain ownership, data contracts, discoverability, interop…
Knowledge Graph in the live execution path of multi-step Large Language Model…
What architectural patterns, operational practices, and failure modes arise when a Knowledge Graph (KG) becomes a key part of the live execution path for multi-step Large Language Model (LLM) systems,…
Hardware load and Large Language Model (LLM) inference performance
How does hardware resource load, Central Processing Unit (CPU), Graphics Processing Unit (GPU), and memory pressure, affect Large Language Model (LLM) inference performance, specifically latency, thro…
Security, Compliance, and Governance Risks of Using Generative AI (GenAI) Tools…
What are the documented security, compliance, and governance risks of using Generative Artificial Intelligence (GenAI) tools such as Microsoft 365 (M365) Copilot for drafting memos, reports, and other…
Implementation Patterns for Regulatory Compliance in Artificial…
What specific implementation patterns, including externalized machine-executable policy rules (Policy-as-Code (PaC)), rules engines, input, tool-use, and output safety controls (guardrails), output va…
Practical Limits of Large Language Model (LLM) Determinism
What are the practical limits of making LLM (Large Language Model)-based decisions or policy enforcement deterministic, even with temperature=0, fixed seeds, and constrained prompts?
Hybrid Architecture Design
How should hybrid architectures be designed so that probabilistic LLMs handle interpretation and insight generation while deterministic layers enforce final governance, compliance, and high-stakes dec…
Governance Policy Application
To what extent must governance policy application be deterministic, consistent, reproducible, and auditable, versus allowing stochastic or probabilistic elements when Artificial Intelligence (AI) or L…
Data Governance Standards and Regulations Applied to Artificial Intelligence…
How do established data governance standards, including International Organization for Standardization and International Electrotechnical Commission (ISO/IEC) 38505, DAMA-DMBOK (Data Management Body o…
Extending Traditional Data Governance Frameworks to Address Large Language…
How can traditional data governance frameworks be extended or mapped to address the inherent non-determinism and uncertainty about whether deployed behavior remains aligned with intended use in modern…
Compliance Risks of Relying on Stochastic Large Language Model (LLM) Outputs…
What evidence or guidance exists on the compliance risks of relying primarily on stochastic Large Language Model (LLM) outputs for governance, privacy, or regulatory decisions?
Orthogonality thesis under modern Large Language Model (LLM) training and…
How should the orthogonality thesis be interpreted for modern Large Language Models (LLMs) given current pre-training and post-training methods, and what does that imply for enterprise risk when agent…
What are the primary behavioural and structural drivers of unsanctioned AI…
What are the primary behavioural and structural drivers of shadow Artificial Intelligence (AI) adoption, meaning unsanctioned use of AI tools without formal approval or oversight, in enterprises after…
What tiered human oversight models maintain meaningful human-in-the-loop (HITL)…
Under high-volume deployment of multi-step Artificial Intelligence (AI) systems, what factors cause human-in-the-loop (HITL) oversight to degrade into rubber-stamping, meaning approval without genuine…
What metrics beyond code acceptance rates best capture net organisational value…
What metrics beyond code acceptance rates and lines of code best capture net organisational value when Artificial Intelligence (AI) coding tools such as GitHub Copilot are adopted with productivity ma…
How do coupled enterprise risks manifest differently in agentic Artificial…synthesis
How do the coupled enterprise risks, capability debt, incentive-driven shadow Artificial Intelligence (AI) adoption, skill decay, and oversight failure, manifest differently in agentic AI, meaning aut…
How can organisational capability debt be rigorously defined and measured as a…
How can capability debt, the accumulated organisational deficit in review quality, judgment, process maturity, and skill inventory, be rigorously defined, measured, and tracked as a leading indicator…
To what degree does over-reliance on AI tools accelerate measurable skill decay…
To what degree and through what mechanisms does over-reliance on Artificial Intelligence (AI) tools, particularly tools that can plan or act across multi-step workflows, accelerate measurable skill de…
Updating the enterprise Artificial Intelligence ecosystem capability reference…synthesis
How should the enterprise Artificial Intelligence (AI) ecosystem capability reference architecture (as expressed in `2026-04-22-enterprise-ai-capability-model`, `2026-05-05-enterprise-ai-capability-st…
Production incidents linked to Artificial Intelligence systems
What documented production incidents over the last five years were caused or materially contributed to by Artificial Intelligence (AI) systems, and what recurring failure modes and mitigations were id…
Integrating 2026-05 security and supply chain findings into the enterprise…synthesis
How should the enterprise Artificial Intelligence (AI) ecosystem capability reference architecture (as expressed in `2026-04-22-enterprise-ai-capability-model` and the `2026-05-05-enterprise-ai-capabi…
What is the architecture and practical applicability of OpenFactCheck as an…
What is the architecture, evaluation methodology, and practical applicability of OpenFactCheck as an automated, modular, claim-level fact-checking pipeline for Artificial Intelligence (AI)-generated c…
How do open-weight policy enforcement reasoning models, exemplified by OpenAI's…
How do open-weight, meaning released-weight and self-hostable, policy enforcement reasoning models, exemplified by OpenAI's gpt-oss-safeguard, classify text against strict, customizable policies, and…
What is the minimal viable schema for an Artificial Intelligence bill of…
What is the minimal viable set of schema properties required to describe Artificial Intelligence (AI) system dependencies for systems that use prompts, retrieval knowledge bases, memory, and tools in…
Why does Software Bill of Materials (SBOM) fail as a complete inventory model…
Why do traditional Software Bill of Materials (SBOM) concepts fail to adequately describe the dependency, provenance, and runtime composition of agentic Artificial Intelligence (AI) systems, and what…
How can a runtime-observed Artificial Intelligence Bill of Materials (AIBOM) be…
How can a dynamic, runtime-observed Artificial Intelligence Bill of Materials (AIBOM) be generated for an agentic Artificial Intelligence (AI) system, capturing execution traces, transient Retrieval-A…
How do you capture a runtime-observed Artificial Intelligence Bill of Materials…
How do you instrument a real agentic Artificial Intelligence workload, meaning a tool-using workload that plans or acts across multiple steps, to capture a runtime-observed Artificial Intelligence Bil…
How does the European Union (EU) AI Act and related international AI governance…
- [fact; source: https://owaspaibom.org/] Artificial Intelligence Bill of Materials (AIBOM) is used here in the Open Worldwide Application Security Project (OWASP) sense of an artifact intended to mak…
What introspection, export, and control surfaces actually exist across…
What logs, traces, audit Application Programming Interfaces (APIs), Artificial Intelligence Bill of Materials (AIBOM) export capabilities, version-pinning mechanisms, allowlists, and policy hooks actu…
How should identity, delegation chains, and permission scopes be formally…
How should identity, delegation, and permission scopes be formally represented in an Artificial Intelligence Bill of Materials (AIBOM) schema to enable end-to-end attribution, "who authorized what", a…
How do OAuth 2.0, OpenID Connect, and SPIFFE token propagation work in real…
How do OAuth 2.0 (Open Authorisation), OpenID Connect (OIDC), and SPIFFE (Secure Production Identity Framework for Everyone) token propagation mechanisms work in real multi-agent Artificial Intelligen…
What security and governance risks can a declared and runtime-observed…synthesis
What categories of security and governance risk can an Artificial Intelligence Bill of Materials (AIBOM), an artifact intended to make artificial intelligence systems transparent, auditable, and secur…
How do you construct a declared design-time Artificial Intelligence Bill of…
How do you extract and construct a declared design-time Artificial Intelligence Bill of Materials (AIBOM), covering model, prompt or system instruction, tools, Retrieval-Augmented Generation (RAG) kno…
What systematic review methodologies and Artificial Intelligence (AI)-assisted…
What systematic review methodologies, Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA), Cochrane review, narrative synthesis, meta-ethnography, and realist synthesis, and wh…
How does STORM's perspective discovery step work, and what is the…
How does the STORM (Synthesis of Topic Outlines through Retrieval and Multi-perspective question generation) system's perspective discovery step generate diverse expert viewpoints before decomposing a…
What structured approaches and Artificial Intelligence (AI) agent workflow…
What structured approaches, from academic writing pedagogy, Artificial Intelligence (AI)-assisted writing tools, and agent workflow design, exist for converting synthesised research findings into poli…
What entity-relation schema and write/query patterns best support cross-session…
What entity-relation schema and write-query prompt patterns best support cross-session research provenance and concept reuse for an Artificial Intelligence (AI) research agent using the `@modelcontext…
What technical architecture best supports cross-item synthesis, knowledge…
What technical architecture best supports three distinct but related capabilities in a file-based research corpus (~200 items, growing weekly): (1) a meta-distillation layer that proactively aggregate…
What automated claim verification approaches against scientific literature…
What automated claim verification approaches against scientific literature, specifically arXiv preprints, are used in research synthesis systems, what search strategies maximise recall and precision f…
What adversarial review and red-teaming methods are most effective for…
What adversarial review and red-teaming methods, drawn from Artificial Intelligence (AI) safety research, debate-based evaluation, formal argumentation theory, and scientific peer review practice, are…
What architectural capabilities and contractual conditions are required to…
What architectural capabilities and contractual conditions are required for an enterprise to maintain multi-platform portability and mitigate Artificial Intelligence (AI) vendor lock-in risk from: Mic…
What capability and control design is needed to mitigate incentive…
What capability and control design is needed, at enterprise scale, to mitigate incentive misalignment (where individuals are rewarded for bypassing governance), shadow Artificial Intelligence (AI) (AI…
How should human-in-the-loop (HITL) design be adapted when AI review volume…
How should human-in-the-loop (HITL) design be adapted when Artificial Intelligence (AI) review volume reaches the point where human reviewers become a throughput bottleneck or default to rubber-stampi…
What security capabilities are required in an enterprise Artificial…
What security capabilities are required in an enterprise Artificial Intelligence (AI) system, beyond basic Application Programming Interface (API) access controls and audit logging, to address prompt…
Vendor-agnostic enterprise Artificial Intelligence (AI) capability model
What is the complete set of architectural capabilities required to run Artificial Intelligence (AI) safely at scale in a regulated enterprise, how do Microsoft's Copilot family (Microsoft 365 Copilot…
What does TerminalBench reveal about minimal toolsets and coding agent…
What does the TerminalBench benchmark reveal about the relationship between toolset minimalism and coding agent performance, and what design principles does it suggest for effective Artificial Intelli…
What principles and governance practices enable sustainable, high-quality…synthesis
What principles and governance practices, spanning harness design, task selection, human oversight, and open-source software (OSS) ecosystem health, enable sustainable, high-quality software developme…
What are the design tradeoffs of self-modifying, malleable Artificial…
What are the design tradeoffs, in capability, reliability, safety, and maintainability, between self-modifying agent architectures, where the agent can alter its own toolset, prompts, or extensions at…
What strategies are effective for open-source software maintainers dealing with…
What strategies are effective for open-source software (OSS) maintainers in filtering, managing, and sustaining project health against a rising volume of low-quality Artificial Intelligence (AI) agent…
What is the evidence for human oversight as an effective quality gate in…
What is the empirical evidence that human oversight, specifically the human bottleneck property of limited throughput and pain response, functions as an effective quality gate, meaning the control poi…
What design patterns govern effective extension and plugin systems for…
What design patterns and architectural principles govern effective extension and plugin systems for Artificial Intelligence (AI) coding agent harnesses, and what are the key trade-offs between extensi…
How do errors compound in Artificial Intelligence (AI)-agent-heavy codebases,…
How do errors ("boooos") compound in codebases developed with high volumes of AI agent-generated code, including how local patches cause global regressions, and what review and governance strategies c…
What are best practices for transparent, user-controlled context management in…
What are the best practices for transparent, deterministic, and user-controlled context management in Large Language Model (LLM) coding agent harnesses, and what are the demonstrable harms of opaque c…
What criteria define tasks where Artificial Intelligence (AI) coding agents…
What empirically grounded criteria define the characteristics of software development tasks where Artificial Intelligence (AI) coding agents reliably add value, versus tasks where agent autonomy intro…
Artificial Intelligence coding harness quality benchmarks
What benchmarks, metrics, and evaluation methodologies are used to measure the quality of Artificial Intelligence (AI) coding harnesses, including Integrated Development Environment (IDE) plugins, age…
Prof Suraj Srinivasan's automation and augmentation scores
What does Prof Suraj Srinivasan's research framework for measuring Automation and Augmentation (A&A) scores across job roles and industries reveal about which roles face full Artificial Intelligence (…
Ubiquitous Language in Artificial Intelligence (AI)-augmented development
How significantly does maintaining a living Ubiquitous Language (UL), in the Domain-Driven Design (DDD) sense of a shared, precise domain vocabulary used consistently in both code and conversation, im…
Test-Driven Development (TDD) and fast feedback loops in Artificial…
How does enforcing Test-Driven Development (TDD) with AI coding assistants, writing failing tests before asking the AI to implement, change the quality and stability of the AI output compared to "writ…
Strategic versus tactical roles in Artificial Intelligence (AI)-augmented…
In an Artificial Intelligence (AI)-augmented software team, what is the optimal division of labour between the human developer, who owns strategic design, interface definition, and architectural overs…
Software Engineering fundamentals and AI code generationsynthesis
Drawing on the planned seven-item research programme on Software Engineering (SE) fundamentals in Artificial Intelligence (AI)-augmented development, six completed primary items plus external anchors…
Grill-Me technique: iterative structured interviewing for human and Artificial…
How effectively does the "Grill Me" technique, relentless iterative structured interviewing of the human developer by the AI assistant to build a shared design concept before generating any code, redu…
Fundamentals-first versus specs-to-code
What empirical patterns emerge when comparing real-world software projects built with a strict fundamentals-first Artificial Intelligence (AI) workflow, structured alignment, modules with simple inter…
Deep modules in AI-augmented development
How much more effective is Artificial Intelligence (AI) at understanding, navigating, and correctly modifying a codebase composed of deep modules with simple interfaces versus one filled with many sha…
Artificial Intelligence code entropy and complexity
Does repeated Artificial Intelligence (AI) code generation without strong architectural guardrails demonstrably increase software entropy and complexity over time, as predicted by the entropy model de…
Anthropic Claude Teams or Enterprise vs Microsoft 365 Copilot Coworksynthesis
How do Anthropic Claude, specifically the Team and Enterprise plans, and Microsoft 365 (M365) Copilot Cowork compare across capability, pricing, user experience, and guardrails, and what are the secur…
The orthogonality thesis in Artificial Intelligence (AI) alignment
What is the orthogonality thesis in Artificial Intelligence (AI) alignment, what is the current evidence for and against it, and what are its practical implications for Explainable Artificial Intellig…
Human cognitive bias toward Artificial Intelligence (AI) correctness and…
To what extent do humans systematically over-trust AI-generated explanations, and what mechanisms, automation bias, RLHF-induced sycophancy in post-training, and the polysemantic nature of internal mo…
Explainable Artificial Intelligence (XAI)
What is the current state of Explainable Artificial Intelligence (XAI) research, who leads it and what are the primary techniques, and how does XAI intersect with regulatory obligations, audit require…
Is knowledge scaffolding an established concept within context engineering for…
Is knowledge scaffolding an established concept within context engineering for Large Language Models (LLMs) and Artificial Intelligence (AI) agents, and if so, how is it defined, implemented, and dist…
Which software categories face declining demand versus increasing demand as…
As Artificial Intelligence (AI) coding agents, such as Anthropic Claude Code, OpenAI Codex, and GitHub Copilot Workspace, make custom software generation materially cheaper, which categories of commer…
Large Language Model (LLM)-as-judge as pipeline validation checkpoints
Which organisations, projects, and frameworks are defining and operationalising Large Language Model (LLM)-as-judge evaluation, the use of one model to assess another model's outputs, as automated val…
Alternative Continuous Integration and Continuous Delivery pipeline platforms…
What alternative Continuous Integration and Continuous Delivery (CI/CD) pipeline platforms, specifically Harness, Amazon Web Services (AWS) CodeBuild and CodeDeploy, and Jenkins, can serve as the gove…
Universal Entity Lifecycle Governance Framework (UELGF) extension
What explicit human oversight and accountability requirements, covering named human owners for every governed entity, defined escalation paths for high-risk autonomous actions, accountability designat…
Universal Entity Lifecycle Governance Framework (UELGF) extension
What agentic Artificial Intelligence (AI)-specific risk categories, specifically emergent behaviour, goal misalignment, multi-agent interaction failures, and hallucinations in decision loops, are insu…
ServiceNow workflow orchestration and agentic Artificial Intelligence (AI)…
What workflow orchestration and governance capabilities does ServiceNow currently provide for Artificial Intelligence (AI) agent workloads, specifically its identity resolution, permissions, audit tra…
Governance-as-moat thesis and prior research implications
How does the thesis advanced in the April 2026 Liam Hyland and Leonis Capital ServiceNow analysis, that governance is the durable, non-replicable value layer in AI-augmented enterprise technology stac…
Universal Entity Lifecycle Governance Framework (UELGF)
What is the complete specification of the Universal Entity Lifecycle Governance Framework (UELGF), integrating foundational definitions and principles, entity taxonomy and Confidentiality, Integrity,…
Universal Entity Lifecycle Governance Framework (UELGF)
What are the foundational definitions, formal principles, and architectural properties required to specify the Universal Entity Lifecycle Governance Framework (UELGF) such that it applies consistently…
Universal Entity Lifecycle Governance Framework (UELGF)
What canonical entity taxonomy and Confidentiality, Integrity, and Availability (CIA) classification system should the UELGF use to determine governance intensity, ensuring that every entity type, fro…
Invariant-based anomaly detection in the Policy Information Point (PIP)
How can the Policy Information Point (PIP) detect when a governed asset's transient operating context is being used, intentionally or through task creep, to suppress or obscure a permanent invariant,…
Universal policy synchronisation and integrity
What mechanism ensures that the Policy Decision Point (PDP) evaluates a governed asset against logically identical policy at every lifecycle phase, such that a soft gate in Development and a hard gate…
Policy Administration Point (PAP) dynamic policy profiling and proportionality
How can a Policy Administration Point (PAP) dynamically map a governed asset's metadata, specifically its invariants and Confidentiality, Integrity, and Availability (CIA) ratings, to a proportional a…
Out-of-band policy invalidation and remediation
What consistency model governs [Policy Administration Point (PAP)](https://docs.oasis-open.org/xacml/3.0/xacml-3.0-core-spec-os-en.html)-to-[Policy Enforcement Point (PEP)](https://docs.oasis-open.org…
Cryptographic preservation and runtime evaluation of original intent
What representation of original intent, captured at the Getting Started phase, is simultaneously cryptographically verifiable and semantically stable enough to function as a meaningful evaluation base…
What is the precise technical distinction between code generation and other…
What is the precise technical distinction between code generation and other Large Language Model (LLM)-generated outputs in terms of external verifiability, specifically, that code operates in a forma…
What is Yann LeCun's complete argument against Large Language Models as a path…
What is Yann LeCun's complete and precise argument against Large Language Models (LLMs) as a path to autonomous machine intelligence, meaning Artificial Intelligence (AI) that can reason, plan, and ac…
What does synthesising LeCun's architectural critique of Large Language Models…
What does the synthesis of Yann LeCun's architectural critique of Large Language Models (LLMs), no causal world model, no consequence reasoning, verifiable only in formal systems, with the systems cap…
What identity and access management model is required for Artificial…
What identity and access management (IAM) model is required for non-human actors, AI agents and low-code artefacts, operating within enterprise systems, specifically: how should machine identities be…
What control-plane architecture is required to manage Artificial Intelligence…
What control-plane architecture is required to manage AI agents and low-code systems as distributed, semi-autonomous actors within enterprise environments, specifically, how should policies be created…
Implicit rate-limiting controls removed by agentic Artificial Intelligence (AI)
Prior to agentic Artificial Intelligence (AI), the blast radius of ungoverned citizen development was implicitly bounded by human speed, attention, fatigue, and working hours, controls that are not do…
Deployment pipeline as the only enforceable control gate for citizen-developed…
In an environment where citizen development tooling is already licensed and accessible to non-technical staff, and where the distinction between personal productivity and production automation has col…
Access control amplification under agentic operations
Agents do not inherit a user's typical behaviour, they inherit the worst-case interpretation of that user's full permission set, because they operate without fatigue, attention limits, or working hour…
Policy coherence as a machine-checkable prerequisite
Contradictory or outdated policy documents are a chronic governance failure that organisations tolerate because the consequences under human operation are slow-moving. Under agentic operation, agents…
Permission-safe Retrieval-Augmented Generation (RAG) in enterprise information…
What are the technical constraints on permission-safe Retrieval-Augmented Generation (RAG) in an enterprise information architecture with incoherent access controls, collaboration groups created ad ho…
Dependency ordering of foundational conditions for safe agentic Artificial…
The foundational conditions for safe agentic AI deployment in a regulated financial institution are not independent, they form a dependency graph in which policy coherence is a prerequisite for inform…
Systems capability debt, citizen development, and agentic AI risk
Does the synthesis of technical debt literature (Cunningham, Kruchten), systems capability research, transaction cost economics (Coase, Williamson), operational risk frameworks (Basel III/IV, Risk and…
Regulatory and standards preconditions for deployment of Artificial…
Under applicable regulatory and standards frameworks, including Australian Prudential Regulation Authority (APRA) CPS 230, the European Union (EU) Digital Operational Resilience Act (DORA), Payment Ca…
Global artificial intelligence agent regulation in financial services
What regulatory obligations do financial-services regulators globally, including the European Union (EU), Australia, New Zealand (NZ), the United States (US), and the United Kingdom (UK), impose on Ar…
Business-led low-code agent governance
Under what conditions does business-led low-code Artificial Intelligence (AI) agent creation produce durable organisational value versus technical debt and governance fragmentation, and what foundatio…
Recall competitive landscape and clone feasibility
What core capabilities does Recall provide, who else is building similar products (including projects in the davidamitchell GitHub organization and relevant open-source tools), what components can we…
Enterprise AI capability model for use-case maturity decisions
What enterprise-wide Artificial Intelligence (AI) capability model best supports deciding whether a candidate AI use case requires net-new foundational capabilities or can reuse capabilities already b…
Harness-level selection and use of tools, agents, skills, prompts, and…
When should teams choose tools, agent definition files, skills, prompts, instruction files, and AGENTS.md, and what verifiable best practices align with how major harnesses actually select and apply e…
Artificial Intelligence (AI)-assisted daily productivity digest
What are the established patterns and tooling approaches for using Artificial Intelligence (AI) to generate actionable daily and weekly productivity digests from personal task management systems?
Sup…
Claude mythos: character, soul documents, and narrative identity in large…
What is the "Claude mythos" - the narrative, character, and values framework Anthropic has built into Claude - and who else in the industry is doing similar work on giving large language models (LLMs)…
AI company hiring strategies
What do recent and historical job advertisements and hiring patterns at major Artificial Intelligence (AI) companies signal about their current and emerging strategic priorities, and where are the mos…
The shape of organisations when software is no longer the constraint
Inside an organisation that requires software to be built, integrated, and maintained (including Commercial Off-The-Shelf (COTS) systems, Software-as-a-Service (SaaS) platforms, and bespoke-built syst…
oh-my-codex and AI Agent Workflow Patterns
What patterns from oh-my-codex (OMX) and similar AI agent workflow projects (AGENTS.md, SKILL.md, etc.) are most applicable to improving the instructions, skills, agents, and tooling across davidamitc…
Anthropic Claude Code leak
What does the accidental March 2026 leak of Anthropic's Claude Code source code reveal about: (1) the codebase architecture, (2) how key engineering problems are solved, (3) the prompting and instruct…
AI Funding and Capital Investment Landscape
Which Artificial Intelligence (AI)-related and tech companies are receiving and deploying capital investment in 2023-2026, who the major investors are, where investment is concentrated, and what actio…
Large Language Models as offensive security tools
What is the current state of Large Language Model (LLM)-driven offensive security capability: can LLMs autonomously discover and exploit zero-day (0-day) vulnerabilities, what does the empirical evide…
The role of AGENTS.md in a repo using .github/copilot-instructions.md as the…
`AGENTS.md` has emerged as the cross-tool convergence format for agent project instructions, supported by OpenAI Codex, GitHub Copilot, Claude Code, Cursor, Aider, Gemini Command Line Interface (CLI),…
Multi-agent repo setup
What are the best practices for setting up a GitHub repository so that it can be worked on effectively by multiple Artificial Intelligence (AI) agents, specifically: (1) Claude via the iOS Claude app,…
Claude Code on the web
Does Claude Code on the web automatically initialise git submodules when cloning a repository, and if so, can it access private submodules (such as `davidamitchell/Skills` referenced at `.github/skill…
Agent instruction loading and skills access
Given the current repo setup -- instructions in `.github/copilot-instructions.md`, skills submodule at `.github/skills/`, no `AGENTS.md` at root, no `CLAUDE.md` at root -- what does each agent actuall…
Customer contact as strategic signal
When customers contact a business, what are they actually seeking — and how should organisations decide between self-service automation and human interaction to maximise long-term customer value?
Cost reduction is not a strategy
Why is cost reduction insufficient as a business strategy, and how does framing artificial intelligence (AI) primarily as a cost-cutting tool risk destroying value through missed opportunities?
Public sentiment on AI in banking and high-trust institutions
What does current (2024–2025) survey data reveal about customer sentiment toward Artificial Intelligence (AI) in banking and high-trust Financial Services (FS) institutions — in Australia, across Asia…
Agent orchestration patterns
What agent orchestration patterns, verification strategies, and multi-model delegation techniques are demonstrated by Burke Holland's Anvil, Max, and the orchestrator/planner/coder/designer multi-agen…
Artificial Intelligence (AI) agents as finishers and synthesisers
What agent configurations, prompt strategies, orchestration patterns, and tooling choices allow an AI agent (or agent team) to act as a reliable *finisher* and *synthesiser* - completing work that a h…
How to best use awesome-copilot in this repo and across personal repos
What resources from `davidamitchell/awesome-copilot` — GitHub Copilot (GHC) instructions, skills, agents, workflows, hooks, and plugins — provide the most leverage when applied to this Research repo a…
Applied context engineering
What practical patterns, workflow best practices, and agent development guidelines emerge from synthesising the `muratcankoylan/Agent-Skills-for-Context-Engineering` skill library with the context eng…
Coding AI Agent Skills Survey
What actively maintained, publicly available agent skills, prompt libraries, instructions files, and system prompts exist — from vendors such as Microsoft and from the Open Source Software (OSS) commu…
Dependency Mapping Across .NET Codebases, Terraform, Dynatrace, Confluence, Log…
What practical tools and methodologies are being used to map dependencies across .NET codebases, Terraform configurations, Dynatrace Application Performance Management (APM) monitoring, and solution d…
More formal proof engineering
What does Leanstral - an open-source agent for formal proof engineering - offer as a practical path to trustworthy, formally verified software built with Artificial Intelligence (AI) assistance, and h…
Explore to exploit: the synthesis step that makes exploitation pay off
When technology is moving fast, what is the synthesis step between exploration and exploitation, why is it so commonly skipped, what are the costs of skipping it, and what strategies allow organisatio…
Application Programming Interface (API) Context Hubs, Retrieval-Augmented…
What approaches are being used to enable Artificial Intelligence (AI) agents to discover, understand, and invoke external Application Programming Interfaces (APIs), and how do the three major emerging…
Stateless-agent assumption failure
When an agentic workflow spans multiple session boundaries — each session starting with a fresh context window and no memory of prior runs — what are the mechanisms by which external state becomes orp…
Vision-Language Joint Embedding Predictive Architecture (VL-JEPA) and concept…
What is Vision-Language Joint Embedding Predictive Architecture (VL-JEPA) - specifically its concept prediction mechanism - and what practical options exist for a developer consumer of existing fronti…
Intent Driven Development
What is Intent Driven Development (IDD) — as a methodology that moves past Test Driven Development (TDD) and Specification Driven Development (SDD) — and what context and concept layering mechanisms a…
GitAgent and declarative agent definition
What is GitAgent (https://github.com/open-gitagent/gitagent), how can it be used in this repository, what concepts does it build on and produce, and how does the broader idea of declarative agent defi…
Adaptive Policy-Based Authorization (APBA)
How does Adaptive Policy-Based Authorization (APBA) align with the dynamic access-control requirements of National Institute of Standards and Technology (NIST) Special Publication (SP) 800-53 and ISO/…
Trusting Trust and AI Corpus Contamination
Ken Thompson's "Trusting Trust" argument shows that you cannot verify a compiler by reading its source code if the compiler was compiled by a compromised toolchain — the contamination lives in the bin…
Prompt injection threat landscape
What is the current state of the prompt injection threat in agentic artificial intelligence (AI) systems: who is exploiting it, who is defending against it, and what does the research community consid…
Aligned Decision-Making
What framework should an organisation adopt to ensure that AI agents making or supporting decisions have access to the right layered organisational context — spanning regulatory boundaries, values and…
Adam Smith, Organisational Design, Desire Paths, and AI Strategy
What can Adam Smith's insights into human nature and morality - drawn from *The Theory of Moral Sentiments* (ToMS) and *The Wealth of Nations* (WoN) - teach us about designing organisations that align…
Reliable Software in the LLM Era
What strategies and formal-methods tooling exist for maintaining software reliability in the Large Language Model (LLM) era, and what does the Quint formal specification language ecosystem - including…
ChatGPT Actions and custom GPTs
Can a ChatGPT custom Generative Pre-trained Transformer (GPT) be configured with Actions that: (a) call a self-hosted HTTP endpoint to add a memory, (b) call `search_brain` before responding to surfac…
SWAT technique in a fresh-context loop
When the SWAT (Strengths, Weaknesses, Assumptions, Threats) technique is executed repeatedly in a loop where each invocation uses a fresh Large Language Model (LLM) context window and the caller blind…
AI inverted the knowledge-work scarcity equation
Before Artificial Intelligence (AI), throughput (volume of output) was the binding constraint on knowledge work. AI has
dramatically reduced the cost of generating output. Does the evidence support th…
Failure mode taxonomy
The five-layer failure mode taxonomy established in `2026-03-10-ai-concept-classification-taxonomy.md` (Q5) provides a structurally sound classification, but leaves three empirical gaps unanswered: (1…
Exploration-synthesis gap
During periods of rapid exploration — such as the current wave of Artificial Intelligence (AI) / Large Language Model (LLM) adoption inside organisations — individuals and teams routinely duplicate ef…
AI amplified the coordination tax
Artificial Intelligence (AI) has increased per-person output by 5–10x. If coordination cost scales with the square of team size
(see `2026-03-12-team-size-limits-brooks-dunbar-network-theory.md`), wha…
Research loop evaluation rubric
What structured rubric should be used to evaluate the outputs of this repository's research loop agent — and what does a minimal viable implementation of a Continuous Integration (CI)-integrated eval…
Superpowers as inspiration
What ideas, patterns, and workflow practices from `davidamitchell/superpowers` (a fork of `obra/superpowers`) can be used as inspiration for improving agent tooling in `davidamitchell/Latest-developme…
Language designed for LLM agents to produce
Is anyone actively developing a programming language or structured output format specifically designed for LLM agents to generate — rather than humans to write — that structurally addresses generation…
Agent evaluation framework
What evaluation framework allows systematic comparison of agent implementations across multiple repositories — identifying what problems each is solving, whether concepts are used idiomatically or in…
Adversarial agents with shared goals
What is the design pattern for a system of agents — human or AI — that share a common goal but deliberately occupy different competency domains and time horizons? How does "adversarial collaboration"…
AI concept classification taxonomy
What is a coherent, internally consistent classification taxonomy for the core concepts in AI-assisted and agentic systems — covering prompt types, instruction types, prompt/content/intent engineering…
Formal intent specification and language choice for AI alignment in agentic…
Can formal specification of task intent structurally eliminate reward hacking and intent mismatch in agentic coding systems? What is the expressiveness-verifiability tradeoff at each level of the spec…
ServiceNow AI: Knowledge Management, RAG Pipelines, and Agent Frameworks
How is ServiceNow evolving its platform to support AI-powered knowledge management, retrieval-augmented generation (RAG), and agent frameworks — and what should an organisation investing in ServiceNow…
ServiceNow Platform Strategy
Given the findings from the Common Service Data Model (CSDM) data modelling, process mapping, and AI capability research, how should an organisation develop a coherent, practical ServiceNow platform s…
AI coding harnesses: agent execution model, memory, and context management…
What are the core architectural and philosophical principles behind the AI coding harnesses (agentic IDEs and agent runtimes) published or released by Anthropic, OpenAI, and the broader ecosystem of c…
Emergent Patterns in Software Engineering Prompts and SDLC Guidance
What are the current and emergent best practices for crafting AI agent prompts and tooling guidance tailored to each phase of the Software Development Life Cycle (SDLC) — covering discovery, requireme…
Claude for iOS: MCP remote integration for memory capture and retrieval
Does the Claude iOS app support MCP connections to a remote server? If so: (a) what transport is supported (HTTP/SSE vs stdio), (b) does it require the same `.mcp.json` config as Claude Desktop, (c) w…
An Integrative Framework for Agent Decision-Making
How can the DIKW (Data → Information → Knowledge → Wisdom) progression be operationalised within agentic systems to produce intent-aligned, context-aware decisions that reconcile conflicting knowledge…
AI capability is not a data problem - why the data/analytics department is the…
What is the strongest case - technical, architectural, organisational, legal, and regulatory - that an organisation's AI capability should NOT be owned by or coupled to its data/analytics department o…
Guiding Headless Agents via LSP-Like Mechanisms for Org Policy Conformance
Who is building solutions that allow headless autonomous coding agents to be guided in real time by LSP-like mechanisms — rather than CI gates or pre-commit hooks — to conform to an organisation's sec…
Self-improving Artificial Intelligence (AI) agent evaluation loop architecture
What is the most principled architecture for a Self-Improving AI Agent Evaluation Loop — specifically, how should a nested inner/outer loop be designed so that a "Meta-Optimizer" rewrites system promp…
Free energy, entropy, and life
Why do living organisms need predictive brains? What is the precise relationship between the thermodynamic concept of entropy (disorder), the information-theoretic concept of free energy (surprise), K…
Artificial Intelligence (AI) coding assistant deployment outcomes
Which organisations have published or disclosed coherent AI strategies specifically targeting software engineering, what outcomes have they measured, and what does the trajectory from AI-assisted codi…
Research output types
What are the possible output types from a research item, and how should each type be handled, stored, and acted upon?
Evaluating and improving autonomous research loop quality
How can the quality of research items produced by the `research-loop.yml` autonomous pipeline be systematically evaluated, and what changes to `research-prompt.md` and the loop's prompting strategy wo…
Knowledge Representation for Agent Context
What techniques — latent semantic extraction, knowledge graphs, concept maps, hierarchical document compression, and layered abstraction — most effectively represent and compress large knowledge corpo…
Artificial Intelligence (AI) agents in financial services line 1 and line 2…
Who is currently building or deploying AI agents specifically positioned to operate within the three lines of defence model — line 1 (business/operational risk management) and line 2 (risk and complia…
AI for Control Testing, Gap Identification, and Policies/Standards Reviews
Which organisations are using AI to automate control testing, identify control gaps, or conduct policies and standards reviews — and what does the current vendor, practitioner, and regulatory landscap…
Transaction Cost Economics
What are the foundational concepts of transaction cost economics (Coase → Williamson → North → Ostrom), and how might the analytical framework map onto software engineering organisation, AI agent desi…
Agent Memory Management and Context Injection
What is the current state of agent memory management systems — beyond RAG — and which approaches best address the real constraints of latency, knowledge freshness, scoping, governance, quality, and di…
Information synthesis
What is the best way to synthesise information from multiple sources in a manner that is minimally lossy — preserving the most signal while compressing volume — drawing on information theory, entropy,…
GitHub Specify, Ralph Loops, and Lisa Planning
What is "Specify" in the context of GitHub-integrated AI development workflows, how does the Ralph loop implement proof-driven development in practice, and what role does Lisa planning play in the spe…
AI Strategy: global and NZ examples, policy frameworks, regulations, and…
What do leading global AI strategies look like, how does New Zealand's regulatory and policy landscape (RBNZ, DIA, MBIE, and others) compare, and what use-case typology — from human augmentation throu…
Predictive processing and active inference
What exactly is the predictive processing / active inference framework, how does it differ from classical feedforward models of perception, and what is the empirical status of the free energy principl…
Jevons Paradox: efficiency gains, demand rebound, and the falling cost of…
How does Jevons Paradox operate across different sectors historically, what are the conditions under which cost or efficiency improvements *do not* increase total demand, and what do current thinkers…