AI & Data Career Foundations · Beginner · ~6 hours
Introduction to Artificial Intelligence
Understand AI fundamentals, generative systems, common applications and responsible-use limits.
Status: Catalogue ready. This lesson version is available in review mode while subject-matter, accessibility and assessment checks are completed.
Understand AI systems and their limits
AI, machine learning and model vocabulary
By the end of this lesson
- Describe AI and machine learning in plain language without treating them as magic.
- Distinguish training data, inputs, outputs, models and evaluation.
- Identify what a simple model can and cannot establish about a person or situation.
What people mean by artificial intelligence
Artificial intelligence is an umbrella term for systems designed to perform tasks that people often associate with human reasoning, perception, language or decision support. The term covers a wide range of approaches, from rules written by people to statistical models that learn patterns from examples. A system can be useful without understanding the world in the way a person does. It can also produce a fluent answer without having reliable knowledge behind that answer. Machine learning is one approach within AI. Instead of specifying every rule, a team gives a model data and an objective so it can find patterns that support a task. A training set is used to fit the model, while validation and test data help evaluate whether it performs beyond the examples it saw. The quality, relevance and representation of the data matter. If historical decisions reflect unfair treatment or missing groups, a model can reproduce those patterns even when the code is technically correct. An input is information supplied to a system. An output is what the system returns, such as a classification, prediction, recommendation or generated text. A model is the learned or programmed mechanism between them. An output is not automatically a fact. It is a result produced under particular data, instructions, settings and evaluation conditions. Keeping these distinctions clear helps a salesperson explain an AI-enabled product without implying certainty or human judgement where it does not exist.
Performance needs a question and a context
There is no single score that tells you whether an AI system is good. Evaluation depends on the task and the cost of error. A model that identifies likely duplicate records may be useful even if a human must review every suggestion. A model used to make an access, credit, health or employment decision needs much stronger evidence, controls and accountability. Accuracy can also hide unequal performance across groups, rare but serious failures or a mismatch between a test set and real use. Ask what the system is intended to support, who is affected, what a wrong output costs and who remains responsible for the decision. Check whether the evaluation data resembles the context in which the system will operate. Look for false positives and false negatives, not only an average score. Test unusual but plausible inputs and record how people can correct an error. If the use case changes, the evaluation may no longer apply. AI literacy includes knowing when not to use AI. A clear rule-based calculation, a human conversation or a small transparent spreadsheet may be safer and easier to inspect. The goal is not to insert AI into every workflow. It is to choose a method that supports the intended outcome while respecting privacy, fairness, accessibility and human agency.
Evidence, judgement and a safe next step
The professional habit in this lesson is to make the decision process visible. Start by naming the AI use case and model output in plain language, then identify the evidence that supports the current view and the evidence that is still missing. Evidence can be a buyer statement, a dated observation, a documented calculation, a permitted source or a clearly labelled hypothesis. It should be possible for another colleague to understand where a conclusion came from without relying on confidence or personal reputation. When the evidence is weak, reduce the strength of the claim rather than decorating it with persuasive language. The next step should be proportionate to what is known. State the task, data, output, error cost and human decision point before discussing a model's benefits. A small, reversible action is often more useful than a grand promise: ask one clarifying question, test one calculation, invite the right stakeholder, or record one assumption for review. Give the other person a genuine opportunity to agree, decline or change the sequence. This matters in commercial work because a fast process is not automatically a good process, and it matters in career development because a polished artefact is not automatically truthful evidence. Keep the boundary explicit. AI literacy must not turn a demonstration, model output or learning exercise into a promise about a person or outcome. Do not put private information, speculative outcomes or unsupported performance claims into a learner portfolio, CRM, dashboard or application. If a decision affects another person, explain who owns it and how they can challenge or correct it. A good submission therefore shows both capability and restraint: it demonstrates what was done, states what was not tested, and identifies a sensible follow-up. Reviewers can reward that clarity even when the proposed answer is incomplete, because responsible professional judgement is part of the skill being developed.
Workplace example
A fictional sales team considers an AI tool that classifies inbound enquiries. The team checks what data trained the classifier, reviews false positives and false negatives, and keeps a human decision-maker for routing. It does not describe the system as understanding a person's intent or as making a final judgement.
Try this walkthrough
- 1.Describe the task, input, output, affected people and responsible decision-maker.
- 2.Separate training data, evaluation data and live-use data in a simple diagram.
- 3.List two ways the model could be wrong and the consequence of each.
- 4.Decide whether AI is proportionate or whether a simpler method is clearer and safer.
Watch for these traps
- Treating a fluent output as proof that a system understands or is correct.
- Using an average accuracy score without asking who experiences errors.
- Assuming historical data is neutral because it comes from a real organisation.
- Adding AI to a workflow without defining human responsibility.
Knowledge check
Choose the response you think is strongest. These checks are practice until the course version has passed subject-matter review; graded attempts will be handled server-side.
For the reviewed pilot, answers are checked on the server. This draft result is not yet added to your official learning record.
Practice studio
Make a decision. Test a response. Leave evidence.
Use the case from this lesson to practise judgement without creating a score or changing your course record.
Scenario decision
A fictional sales team considers an AI tool that classifies inbound enquiries. The team checks what data trained the classifier, reviews false positives and false negatives, and…
Choose a response
Branching roleplay
A colleague asks you to move this situation forward. What do you say next?
Evidence workspace
Draft a short known / assumed / next note from the fictional case. Use fictional details only.
This is a session-only practice draft. It is not saved, graded or added to your learning record. Never enter real personal or confidential information.
Practical assignment
Create a one-page concept map for a fictional enquiry-classification model. Include inputs, output, training and evaluation questions, error types, affected people and the human decision point.
Pass this lesson's knowledge check before recording it as complete.
Portfolio evidence
Practical portfolio project
Compare several fictional AI use cases, identify where AI adds value and document limitations, risks and human oversight needs.
Turn the practical project into a bounded evidence record. Save a draft while you work, then submit it for human review when it is ready.
Sign in to save practical evidence. Sign in