AI Concepts Without the Hype
Look past product labels and classify systems from observable facts, methods, controls, and evidence gaps.
What you will practise
- Describe a system through its objective, inputs, mechanism, outputs, and influence.
- Distinguish fixed rules, machine learning, deep learning, generative AI, retrieval, tool-using assistants, and composite or agentic systems.
- Separate a model from the complete application around it.
- Distinguish training, evaluation, deployment, and inference.
- Mark a classification as uncertain and name the missing evidence.
Module sequence
- Welcome to Module 1: Look Past the LabelWhy labels are weak evidence and what you will learn.
- What Makes a System AI?Describe goals, inputs, methods, outputs, and effects.
- Rules, Machine Learning, Generative AI, and AgentsCompare rules, machine learning, generation, retrieval, tools, and agents.
- Models, Data, Training, and UseSeparate training, evaluation, deployment, and inference.
- Classify Systems from FactsApply TRACE when evidence is incomplete.
- Worked Example: A Policy AssistantMap the parts and controls in a policy assistant.
- Applied checkpointClassify eight made-up systems and review model answers.
- Knowledge checkCheck your understanding with ten retryable questions.
- Module summaryConsolidate the module and prepare for Module 2.
Before you continue
Use only the made-up examples supplied by the course. Do not enter real personal, confidential, client, workplace, health, financial, authentication, or security-sensitive information.
Open the AI-system classification checklist and keep it available for the applied checkpoint.
