中文Start here
Programme / Artificial intelligence

Learn what AI can do, what it cannot know and who remains responsible.

Learners examine data and model behaviour before using AI to make or automate anything. Fluency includes questioning outputs, not merely producing them.

Choose an AI pathway Read the AI literacy guide
Learners can begin with AI literacy before moving into code and model experiments.
OPENQUESTION
01 / Question02 / Understand03 / Build04 / Verify
01
BIAA / FIT

Who this learning direction is for

Placement is based on readiness, prior experience and learning goals rather than age alone.

  • 01Learners curious about how machine-made predictions and generated media work
  • 02Coders or makers ready to investigate data-driven behaviour
  • 03Prerequisite: Introductory AI literacy requires no programming
  • 04Prerequisite: Model-building projects benefit from basic data handling and coding
02
BIAA / QUESTIONS

Questions learners investigate

Each unit begins with a question that can be explored through observation, code, construction or evidence.

  • 01How does training data influence a model’s output?
  • 02What counts as useful evidence that an AI system works?
  • 03When should a person review, override or reject an automated result?
03
BIAA / LEARNING

What learners practise

Technical knowledge is developed alongside planning, documentation, testing and explanation.

  • 01Data collection, labelling and representation
  • 02Classification, prediction and generative systems
  • 03Accuracy, uncertainty and comparison baselines
  • 04Bias, privacy, authorship and environmental cost
  • 05Human oversight and communication of limitations
04
BIAA / OUTCOMES

What learners produce

The expected outcome is a documented learning artefact, not an unverified promise of awards or certification.

  • 01A small, bounded AI investigation or prototype
  • 02A model card or system note describing data, intended use and limitations
  • 03A comparison of outputs under changed prompts, examples or test data
05
BIAA / PROCESS

Learning process and assessment

Learners move through a repeatable cycle and receive feedback against visible criteria.

  • 01Frame a decision or creative task
  • 02Inspect and prepare a suitable dataset or examples
  • 03Build or configure a bounded model interaction
  • 04Test failure cases and document human review
  • 05Assessment: Claims match the test evidence
  • 06Assessment: Data choices and limitations are disclosed
  • 07Assessment: The learner can explain when the system should not be used
06
BIAA / RESPONSIBILITY

Safety, ethics and the next step

Responsible making is part of the curriculum whenever tools, data, AI or autonomous systems are involved.

  • 01Do not upload sensitive, private or copyrighted material without permission
  • 02Generated outputs require verification and clear attribution
  • 03High-impact decisions must not be delegated to a classroom prototype
  • 04Next: Apply AI in a carefully scoped robotics or IoT project
  • 05Next: Continue into AI for Good or research and innovation
BIAA / PROGRAMME FRAMEWORK

What learners make visible

01

Who it is for

  • Learners curious about how machine-made predictions and generated media work
  • Coders or makers ready to investigate data-driven behaviour
02

Starting point

  • Introductory AI literacy requires no programming
  • Model-building projects benefit from basic data handling and coding
03

Core questions

  • How does training data influence a model’s output?
  • What counts as useful evidence that an AI system works?
  • When should a person review, override or reject an automated result?
04

What learners investigate

  • Data collection, labelling and representation
  • Classification, prediction and generative systems
  • Accuracy, uncertainty and comparison baselines
  • Bias, privacy, authorship and environmental cost
  • Human oversight and communication of limitations
05

Project evidence

  • A small, bounded AI investigation or prototype
  • A model card or system note describing data, intended use and limitations
  • A comparison of outputs under changed prompts, examples or test data
06

Learning process

  • Frame a decision or creative task
  • Inspect and prepare a suitable dataset or examples
  • Build or configure a bounded model interaction
  • Test failure cases and document human review
07

How progress is reviewed

  • Claims match the test evidence
  • Data choices and limitations are disclosed
  • The learner can explain when the system should not be used
08

Safety & ethics

  • Do not upload sensitive, private or copyrighted material without permission
  • Generated outputs require verification and clear attribution
  • High-impact decisions must not be delegated to a classroom prototype
09

Where this can lead

  • Apply AI in a carefully scoped robotics or IoT project
  • Continue into AI for Good or research and innovation

Questions to ask

Is this a prompt-writing course?

Prompting may be examined, but the programme also covers data, model behaviour, evaluation, bias, privacy and human responsibility.

Will learners build a production AI system?

No such claim is made. Projects are bounded educational investigations or prototypes designed for supervised learning.

BIAA / NEXT

Continue exploring

01Programme / Coding

Write code that another person can read, test and improve.

Coding at BIAA is a way to model decisions, automate repeatable work and communicate precise instructions—not an exercise in memorising syntax.

02Programme / Research and innovation

A strong project begins by making uncertainty explicit.

This direction helps learners move from a broad interest to a bounded question, defensible method and transparent account of what was—and was not—found.

03Challenge / AI for Good

A good intention does not remove the need to test harm, bias and failure.

Teams propose a bounded use of AI, examine who is affected and test whether a simpler non-AI approach may serve the need better.

04AI literacy guide

AI literacy means understanding a system well enough to question its output.

Young people need more than prompt tricks. They need a working model of data, models, uncertainty, evaluation, attribution, privacy and the decisions that should remain human.

BIAA / NEXT STEP

Choose an AI pathway

Learners can begin with AI literacy before moving into code and model experiments.

Choose an AI pathway