AI InitiativeA connected body of research and engineering: neural learning, memory, mathematical models, coordinated AI work and the environments where those ideas are tested.
The mathematical foundation: representations, parameters, objectives and optimisation. The collection connects foundational models to concrete experiments, with critical reviews beside the original claims.
Explore this theme →26 connected reading items
Research into bounded modules, shared representations, selective training and explicit handoffs between capabilities.
Explore this theme →16 connected reading items
Quantum computing research and quantum-inspired classical methods, examined through retrieval, representation and memory experiments.
Explore this theme →28 connected reading items
Outcome learning, adaptation, evaluation and transfer across models, agents and workflows.
Explore this theme →67 connected reading items
Methods for decomposing goals, directing agents, checking results and retaining lessons that another worker can use.
Explore this theme →82 connected reading items
Coordination across agents, roles, tools and durable execution, with clear responsibility and recovery boundaries.
Explore this theme →45 connected reading items
Cortex and related research into recall, provenance, memory care, scope, histories and reusable knowledge.
Explore this theme →32 connected reading items
Research and implementation around multi-model coordination, entropy, synchronisation and other collaboration signals.
Explore this theme →12 connected reading items
Graph-based analysis of code and workflows, mathematical optimisation, composition and structural reasoning.
Explore this theme →10 connected reading items
Exploration, personas, idea lineage, selection, trust protocols and the governance of a shared intellectual space.
Explore this theme →149 connected reading items
Games as experiences for people and bounded environments for testing agents, learning and decision-making.
Explore this theme →27 connected reading items
Spatial collaboration, product exploration and applications that connect the research to usable experiences.
Explore this theme →8 connected reading items
Neural methods, adaptation and experiments ask how capabilities can improve.
Memory and mathematical structure help connect experience with a new problem.
Workflows and orchestration connect roles, tools, decisions and review.
Games, simulations and shared spaces expose consequences and new questions.
This is the research relationship between the themes, not a claim that every project is already integrated.
Keep the existing explanations, or explore context, containers and components in the new C3 alternatives. Every view retains its source status and implementation limits.
Open the architecture comparison →
Work custody, agent execution, durable scheduling and deployment boundaries.
Memory layers, retrieval, provenance, care and learning boundaries.
Idea exploration, selection, HERALD research and eight guided reading paths.
Structural intelligence, mathematical analysis and graph-based composition.
Mathematical signals, multi-model collaboration and critical evidence review.
Private review edition. The full source collection is available here so we can choose the public presentation. Historical manuscripts, proposed designs and current source reviews retain their own evidence status.