A module on artificial intelligence in practice: rapid prototyping with tools (“vibecoding”), creative expression with generative models and (equally important) responsible, safe and inclusive use of AI. The Association works on promoting the responsible use of technology and on developing knowledge of artificial intelligence.
Key facts
- Age:
- Artificial int.: For younger and older primary-school pupils, and secondary-school pupils
- Data literacy: For younger and older primary-school pupils, and secondary-school pupils
- You and the machine: older primary-school pupils and secondary-school pupils
- Focus:
- Artificial int.: use, literacy, training and deploy
- Data literacy: graphs, statistics, one’s own data
- You and the machine: attention, scams, money, subscriptions, rights
- Prerequisite:
- Artificial int.: Digital literacy; complements Algorithmic programming
- Data literacy: Digital literacy
- You and the machine: Digital literacy
- Ministry of Science and Education (MZO) domain: A, B, D
- Infokup: Informatics fundamentals and DK
- Scope: 3 sets of learning outcomes from the curriculum, 42 topics
- Progression: groups by level of knowledge, not by age
Module aims
Artificial intelligence
- use AI and catch it in a mistake
- understand bias and limits
- (advanced) train and publish a model
Data literacy
- read a graph and notice when it lies
- tell average/median and cause/correlation apart
- make an honest graph from one’s own data
You and the machine
- catch hooks and dark patterns
- recognise scams
- track costs and subscriptions
- use rights over data
Learning outcomes
Artificial intelligence. The participant trains a model on their own data, evaluates it fairly and publishes it as an application or on a device.
Data literacy. The participant reads an unfamiliar graph, corrects a misleading graph and makes an honest story from their own data.
You and the machine. The participant recognises dark patterns and scams, adds up subscriptions and carries out a family security check.
MZO outcomes (NN 22/2018)
Outcomes are mapped to the Informatics curriculum from the Official Gazette (Narodne novine) 22/2018. Codes are unique only within a track, so the track is stated with each code (OŠ – primary school, gimn – grammar school, PMG – science and mathematics grammar school).
- Artificial intelligence: A.7.3 (OŠ); D.2.2 (gimn). AI is not explicit in NN 22/2018 (2018).
- Data literacy: A.7.3, A.8.1 (OŠ); C.2.3 (gimn)
- You and the machine: D.2.4, D.6.1 (OŠ); D.2.2 (gimn)
Levels of mastery
Four levels describe what mastering the module means in practice. A participant progresses by level, not by age.
Artificial intelligence
- satisfactory: uses an AI tool and spots a mistake
- good: splits data into train/test and measures accuracy
- very good: trains a model and evaluates it fairly
- exceptional: publishes a model and explains bias, limits and ethics
Data literacy
- satisfactory: reads a simple graph
- good: notices when a graph misleads
- very good: distinguishes average/median and cause/correlation
- exceptional: builds an honest story from their own data (sampling, checking)
You and the machine
- satisfactory: recognises one ‘hook’ in an application
- good: spots a scam and a dark pattern
- very good: adds up subscriptions and tracks costs
- exceptional: carries out a family check and explains rights over data
Breakdown
The breakdown is organised by sets of learning outcomes and levels of knowledge. A participant enrols at the level they have reached, regardless of age. Click a topic for an explanation of what it covers.
Artificial intelligence
21 topics · For younger and older primary-school pupils, and secondary-school pupils
Foundation level
what a model learns
how a model ‘learns’ from examples (intuition)
useful application
practical and honest use of AI tools
bias and limits
where AI goes wrong and why (bias)
ethics and authorship
rights, sources and honesty when using AI
one generative project
one’s own small project with generative AI
Intermediate level
train/test split
separating data for learning and for checking
accuracy and overfitting
measuring success and the danger of overfitting
confusion matrix
confusion matrix – where the model goes wrong
a ready-made model in code
using an already finished model from code
sensor ML (TinyML)
machine learning from sensors on a small device (TinyML)
data flow (pipeline)
cleaning → features → model → evaluation
Advanced level
neural networks (basics)
perceptron, layers, weights and activations
large language models and prompting
how LLMs work; tokens, context, prompt design
RAG (retrieval + generation)
combining retrieval and generation for more accurate answers
embeddings and vector search
representing meaning with vectors and semantic search
transfer learning
adapting an already trained model
fine-tuning (basics)
further training of a finished model for one’s own task
introduction to reinforcement learning (RL)
an agent learns through reward and trial and error
AI ethics and safety
bias, harms, limits and safe use of AI
fair evaluation
measuring model quality correctly
full deployment (deploy)
publishing a model as an application or on a device
Data literacy
13 topics · For younger and older primary-school pupils, and secondary-school pupils
Foundation level
reading a graph
how to read a graph correctly
how graphs lie
misleading axes, scales and choice of display
distributions
how values are spread (histograms, spread)
Intermediate level
average vs median
when the average misleads and the median describes better
cause ≠ correlation
a connection is not the same as a cause
percentages and risk
how not to be misled by percentages
sampling and bias
random vs convenience sample; selection bias
Advanced level
one’s own data set
collecting one’s own data
cleaning data
tidying messy data
from spreadsheets to pandas
from a spreadsheet to data processing in pandas
hypothesis-testing intuition
null model, simulation and p-value (intuitive)
data provenance and ethics
where data come from, licences, privacy and consent
an honest and clear graph
making a graph that does not mislead
You and the machine
8 topics · older primary-school pupils and secondary-school pupils
Foundation level
attention and ‘hooks’
how applications hold attention (dark patterns)
the screen and one’s own time
managing time in front of a screen
recognising scams
spotting common online scams
Intermediate level
money and cards
safe payment and protecting cards
subscriptions that do not stop
hidden subscriptions that are hard to cancel
buying on the internet
safe and sensible online shopping
Advanced level
the small print and contracts
what the terms you accept actually say
rights over data
your rights (GDPR): access, correction, deletion