Generation and context
Next-token prediction, probabilistic output, and context as the condition under which generation occurs.
AI fundamentals
Why do the observed outcomes occur?
Connect classroom observations with concepts such as tokens, context, sampling and retrieval. Check where an explanation fits and where uncertainty remains.
Model behaviour and variability
Tokens, tokenisation, attention, sampling, parameters, repeated runs, and observable variability.
Grounding and evidence
Embeddings, retrieval, source selection, context insertion, claim support, provenance, and bias.
Process and co-construction
Decomposition, intermediate artefacts, checkpoints, perspectives and roles; PDCA connects planning, execution, evidence-based evaluation and decisions. Project folders, Markdown context, accepted state and reusable skills support continuity between contributions.
Critique and verification
Criteria, reflection workflows, evidence for human judgement, LLM-as-judge, revision decisions, and verification.
Decision support and legitimacy
The difference between fluent generation and legitimate decision support, including calibrated reliance and substantive oversight.
Agentic workflows
Perception, reasoning, planning, capabilities, workflow state, validation, permissions, and control points.
Security and trust boundaries
Instruction/data boundaries, direct and indirect prompt injection, source and memory poisoning, sensitive-information handling, least privilege and agency, layered safeguards, human reliance, and residual risk.