Free public resource

AI glossary for families, teachers, and curious students

Age-appropriate explanations for the vocabulary behind modern AI, from simple smart-machine ideas to models, bias, tokens, tools, and agents.

Start a family discussion

Friendly words for noticing, predicting, privacy, fairness, and trusted adults.

Plan a classroom warmup

Concepts that turn into quick activities, examples, charts, sorting, and debugging.

Explain modern AI

A short path from models and data to prompts, context, tools, and agents.

Browse by first introduction

133 AI literacy terms, spiraled across grade bands

Each term appears where it is first introduced, with chips showing where it returns later. Some terms include shorter or deeper explanations by age band.

K-2

Introduced in K-2

Introduced K-2 · Unit 1

decide

To decide is to choose what to do next based on what you noticed. The camera noticed someone at the door, so it decided to send an alert. Every smart machine runs on notice, then decide.

K-26-89-12
Age-specific explanations
K-2

To decide is to choose what to do next based on what you noticed. The camera noticed someone at the door, so it decided to send an alert. Every smart machine runs on notice, then decide.

6-8

To decide is to choose what to do next based on what you noticed. The camera noticed someone at the door, so it decided to send an alert. Every smart machine runs on notice, then decide.

9-12

To decide is to choose what to do next based on what you noticed. The camera noticed someone at the door, so it decided to send an alert. Every smart machine runs on notice, then decide.

Introduced K-2 · Unit 1

machine

A machine is a tool people build to do a job. A toaster heats bread. A bike gets you somewhere. It does not notice the world or choose what to do next on its own.

K-23-56-89-12
Age-specific explanations
K-2

A machine is a tool people build to do a job. A toaster heats bread. A bike gets you somewhere. It does not notice the world or choose what to do next on its own.

3-5

A machine is a tool people build to do a job. A toaster heats bread. A bike gets you somewhere. It does not notice the world or choose what to do next on its own.

6-8

A machine is a tool people build to do a job. A toaster heats bread. A bike gets you somewhere. It does not notice the world or choose what to do next on its own.

9-12

A machine is a tool people build to do a job. A toaster heats bread. A bike gets you somewhere. It does not notice the world or choose what to do next on its own.

Introduced K-2 · Unit 1

notice

To notice means to pick up information from the world — to see, hear, or sense something. A camera notices motion. A microphone notices sound. Noticing is the first step before any decision.

K-26-89-12
Age-specific explanations
K-2

To notice means to pick up information from the world — to see, hear, or sense something. A camera notices motion. A microphone notices sound. Noticing is the first step before any decision.

6-8

To notice means to pick up information from the world — to see, hear, or sense something. A camera notices motion. A microphone notices sound. Noticing is the first step before any decision.

9-12

To notice means to pick up information from the world — to see, hear, or sense something. A camera notices motion. A microphone notices sound. Noticing is the first step before any decision.

Introduced K-2 · Unit 1

sensor

A sensor is the part of a device that notices the world. Cameras see. Microphones hear. Thermometers feel temperature. Without sensors, a machine has nothing to react to.

K-23-56-89-12
Age-specific explanations
K-2

A sensor is the part of a device that notices the world. Cameras see. Microphones hear. Thermometers feel temperature. Without sensors, a machine has nothing to react to.

3-5

A sensor is the part of a device that notices the world. Cameras see. Microphones hear. Thermometers feel temperature. Without sensors, a machine has nothing to react to.

6-8

A sensor is the part of a device that notices the world. Cameras see. Microphones hear. Thermometers feel temperature. Without sensors, a machine has nothing to react to.

9-12

A sensor is the part of a device that notices the world. Cameras see. Microphones hear. Thermometers feel temperature. Without sensors, a machine has nothing to react to.

Introduced K-2 · Unit 1

smart machine

A smart machine notices something about the world and decides what to do next. A doorbell camera sees a person and chooses to alert your phone. That notice-and-decide loop is what makes it smart.

K-26-89-12
Age-specific explanations
K-2

A smart machine notices something and decides what to do. A doorbell camera sees you and rings. Notice, then decide.

6-8

A smart machine senses input, reasons with rules or a learned model, and acts. That sense-reason-act loop is what separates it from a toaster.

9-12

An agent perceives context, decides among actions, and acts — often with tools in a loop. Smart machine is the kid name; agent is the systems name.

Introduced K-2 · Unit 2

broken

A pattern is broken when it stops repeating the way it should. Red, blue, red, red breaks the pattern. Broken patterns lead to wrong guesses — for you and for machines.

K-23-5
Age-specific explanations
K-2

A pattern is broken when it stops repeating the way it should. Red, blue, red, red breaks the pattern. Broken patterns lead to wrong guesses — for you and for machines.

3-5

A pattern is broken when it stops repeating the way it should. Red, blue, red, red breaks the pattern. Broken patterns lead to wrong guesses — for you and for machines.

Introduced K-2 · Unit 2

chart

A chart is a picture of data you collected — counts, categories, or trends. Charts make patterns visible. When you can see the pattern, you can predict from it.

K-23-5
Age-specific explanations
K-2

A chart is a picture of data you collected — counts, categories, or trends. Charts make patterns visible. When you can see the pattern, you can predict from it.

3-5

A chart is a picture of data you collected — counts, categories, or trends. Charts make patterns visible. When you can see the pattern, you can predict from it.

Introduced K-2 · Unit 2

pattern

A pattern is something that repeats in a predictable way. Red, blue, red, blue is a pattern. When you spot a pattern, you can guess what comes next — and so can a computer.

K-23-56-89-12
Age-specific explanations
K-2

A pattern is something that repeats in a predictable way. Red, blue, red, blue is a pattern. When you spot a pattern, you can guess what comes next — and so can a computer.

3-5

A pattern is something that repeats in a predictable way. Red, blue, red, blue is a pattern. When you spot a pattern, you can guess what comes next — and so can a computer.

6-8

A pattern is something that repeats in a predictable way. Red, blue, red, blue is a pattern. When you spot a pattern, you can guess what comes next — and so can a computer.

9-12

A pattern is something that repeats in a predictable way. Red, blue, red, blue is a pattern. When you spot a pattern, you can guess what comes next — and so can a computer.

Introduced K-2 · Unit 2

predict

To predict is to make an educated guess about what comes next. If most days were sunny this month, you might predict sun tomorrow. Machines predict by finding patterns in past examples.

K-23-56-89-12
Age-specific explanations
K-2

Predict means guess what comes next. Red, blue, red, blue — you predict red. Machines predict by finding patterns.

3-5

To predict is to make an educated guess about what comes next. If most days were sunny this month, you might predict sun tomorrow. Machines predict by finding patterns in past examples.

6-8

To predict is to make an educated guess about what comes next. If most days were sunny this month, you might predict sun tomorrow. Machines predict by finding patterns in past examples.

9-12

Prediction at scale is inference: the model outputs a probability distribution over outcomes given input, not a single guaranteed answer.

Introduced K-2 · Unit 2

repeat

Repeat means something happens again the same way. Patterns work because parts of them repeat. Computers look for repeats in data to find structure.

K-23-56-8
Age-specific explanations
K-2

Repeat means something happens again the same way. Patterns work because parts of them repeat. Computers look for repeats in data to find structure.

3-5

Repeat means something happens again the same way. Patterns work because parts of them repeat. Computers look for repeats in data to find structure.

6-8

Repeat means something happens again the same way. Patterns work because parts of them repeat. Computers look for repeats in data to find structure.

Introduced K-2 · Unit 3

bad examples

Bad examples are incomplete, misleading, or too narrow. Only tiny cats in training means the machine fails on big ones. Bad examples do not break the machine — they teach it the wrong pattern.

K-23-56-8
Age-specific explanations
K-2

Bad examples are incomplete, misleading, or too narrow. Only tiny cats in training means the machine fails on big ones. Bad examples do not break the machine — they teach it the wrong pattern.

3-5

Bad examples are incomplete, misleading, or too narrow. Only tiny cats in training means the machine fails on big ones. Bad examples do not break the machine — they teach it the wrong pattern.

6-8

Bad examples are incomplete, misleading, or too narrow. Only tiny cats in training means the machine fails on big ones. Bad examples do not break the machine — they teach it the wrong pattern.

Introduced K-2 · Unit 3

example

An example is one item you show a machine to teach it. One labeled photo of a cat is an example. Machines learn from many examples, not from a single rule written by hand.

K-23-56-89-12
Age-specific explanations
K-2

An example is one thing you show the machine to teach it. One cat picture is one example.

3-5

An example is one item you show a machine to teach it. One labeled photo of a cat is an example. Machines learn from many examples, not from a single rule written by hand.

6-8

An example is one item you show a machine to teach it. One labeled photo of a cat is an example. Machines learn from many examples, not from a single rule written by hand.

9-12

An example is one row in a labeled dataset — input paired with ground truth. Quality beats quantity when examples are misleading or narrow.

Introduced K-2 · Unit 3

good examples

Good examples cover the kinds of things the machine will actually see later. Big cats and small cats, different colors and angles. Good examples help the machine work on new cases, not just the ones it already saw.

K-23-56-89-12
Age-specific explanations
K-2

Good examples cover the kinds of things the machine will actually see later. Big cats and small cats, different colors and angles. Good examples help the machine work on new cases, not just the ones it already saw.

3-5

Good examples cover the kinds of things the machine will actually see later. Big cats and small cats, different colors and angles. Good examples help the machine work on new cases, not just the ones it already saw.

6-8

Good examples cover the kinds of things the machine will actually see later. Big cats and small cats, different colors and angles. Good examples help the machine work on new cases, not just the ones it already saw.

9-12

Good examples cover the kinds of things the machine will actually see later. Big cats and small cats, different colors and angles. Good examples help the machine work on new cases, not just the ones it already saw.

Introduced K-2 · Unit 3

learn

When a machine learns, it finds what its examples have in common. Show it many cat photos and it picks up shared features. It does not understand cats the way you do — it detects patterns in the data.

K-23-56-89-12
Age-specific explanations
K-2

When a machine learns, it finds what its examples have in common. Show it many cat photos and it picks up shared features. It does not understand cats the way you do — it detects patterns in the data.

3-5

When a machine learns, it finds what its examples have in common. Show it many cat photos and it picks up shared features. It does not understand cats the way you do — it detects patterns in the data.

6-8

When a machine learns, it finds what its examples have in common. Show it many cat photos and it picks up shared features. It does not understand cats the way you do — it detects patterns in the data.

9-12

When a machine learns, it finds what its examples have in common. Show it many cat photos and it picks up shared features. It does not understand cats the way you do — it detects patterns in the data.

Introduced K-2 · Unit 3

train

To train a machine means to show it many examples so it can learn a pattern. You are not typing instructions — you are feeding it data. What it learns depends entirely on what you show it.

K-23-56-89-12
Age-specific explanations
K-2

Train means show the machine lots of examples so it can learn.

3-5

To train a machine means to show it many examples so it can learn a pattern. You are not typing instructions — you are feeding it data. What it learns depends entirely on what you show it.

6-8

To train a machine means to show it many examples so it can learn a pattern. You are not typing instructions — you are feeding it data. What it learns depends entirely on what you show it.

9-12

Training optimizes model weights against a loss function on a dataset — supervised when labels exist, other modes when they do not.

Introduced K-2 · Unit 3

wrong guess

A wrong guess is when the machine picks the incorrect answer on something new. Usually the training examples did not cover that case. Fix the examples, not by telling the machine to try harder.

K-23-56-8
Age-specific explanations
K-2

A wrong guess is when the machine picks the incorrect answer on something new. Usually the training examples did not cover that case. Fix the examples, not by telling the machine to try harder.

3-5

A wrong guess is when the machine picks the incorrect answer on something new. Usually the training examples did not cover that case. Fix the examples, not by telling the machine to try harder.

6-8

A wrong guess is when the machine picks the incorrect answer on something new. Usually the training examples did not cover that case. Fix the examples, not by telling the machine to try harder.

Introduced K-2 · Unit 4

fair

Fair means the system works well for everyone, not just one group. People make systems fair by choosing inclusive training data and checking results across groups. A machine has no feelings — fairness is a human responsibility.

K-23-56-89-12
Age-specific explanations
K-2

Fair means it works well for everyone, not just some people. People pick fair examples — the machine does not choose.

3-5

Fair means the system works well for everyone, not just one group. People make systems fair by choosing inclusive training data and checking results across groups. A machine has no feelings — fairness is a human responsibility.

6-8

Fair means the system works well for everyone, not just one group. People make systems fair by choosing inclusive training data and checking results across groups. A machine has no feelings — fairness is a human responsibility.

9-12

Fairness requires disaggregated metrics across groups. Equal overall accuracy can hide harm to a minority — people set the bar, models report numbers.

Introduced K-2 · Unit 4

honest

Honest means telling the truth about how you made something. If a machine helped you write or draw, say so. Giving credit is not weakness — it is how we stay trustworthy.

K-23-59-12
Age-specific explanations
K-2

Honest means telling the truth about how you made something. If a machine helped you write or draw, say so. Giving credit is not weakness — it is how we stay trustworthy.

3-5

Honest means telling the truth about how you made something. If a machine helped you write or draw, say so. Giving credit is not weakness — it is how we stay trustworthy.

9-12

Honest means telling the truth about how you made something. If a machine helped you write or draw, say so. Giving credit is not weakness — it is how we stay trustworthy.

Introduced K-2 · Unit 4

private

Private information is something only you and trusted adults should know — passwords, home address, personal messages. Smart machines and apps can collect private data, so you decide what to share.

K-26-89-12
Age-specific explanations
K-2

Private information is something only you and trusted adults should know — passwords, home address, personal messages. Smart machines and apps can collect private data, so you decide what to share.

6-8

Private information is something only you and trusted adults should know — passwords, home address, personal messages. Smart machines and apps can collect private data, so you decide what to share.

9-12

Private information is something only you and trusted adults should know — passwords, home address, personal messages. Smart machines and apps can collect private data, so you decide what to share.

Introduced K-2 · Unit 4

safe

Safe means protecting yourself and your private information online. Do not share passwords. Think before you post. If something feels wrong, stop and ask a trusted adult.

K-26-89-12
Age-specific explanations
K-2

Safe means protecting yourself and your private information online. Do not share passwords. Think before you post. If something feels wrong, stop and ask a trusted adult.

6-8

Safe means protecting yourself and your private information online. Do not share passwords. Think before you post. If something feels wrong, stop and ask a trusted adult.

9-12

Safe means protecting yourself and your private information online. Do not share passwords. Think before you post. If something feels wrong, stop and ask a trusted adult.

Introduced K-2 · Unit 4

trusted adult

A trusted adult is a grown-up who helps you make good choices — a parent, guardian, teacher, or counselor. When a machine or message makes you uncomfortable, ask a trusted adult.

K-2

3-5

Introduced in 3-5

Introduced 3-5 · Unit 1

ASCII

ASCII is a standard code that maps letters and symbols to binary numbers. Every keyboard letter has a fixed binary pattern. That is how computers store and send text.

3-56-89-12
Age-specific explanations
3-5

ASCII is a standard code that maps letters and symbols to binary numbers. Every keyboard letter has a fixed binary pattern. That is how computers store and send text.

6-8

ASCII is a standard code that maps letters and symbols to binary numbers. Every keyboard letter has a fixed binary pattern. That is how computers store and send text.

9-12

ASCII is a standard code that maps letters and symbols to binary numbers. Every keyboard letter has a fixed binary pattern. That is how computers store and send text.

Introduced 3-5 · Unit 1

binary

Binary is a way to write information using only ones and zeros — on and off. The letter A, a pixel color, a sound sample — all become long strings of binary inside a computer.

3-56-89-12
Age-specific explanations
3-5

Binary is a way to write information using only ones and zeros — on and off. The letter A, a pixel color, a sound sample — all become long strings of binary inside a computer.

6-8

Binary is a way to write information using only ones and zeros — on and off. The letter A, a pixel color, a sound sample — all become long strings of binary inside a computer.

9-12

Binary is a way to write information using only ones and zeros — on and off. The letter A, a pixel color, a sound sample — all become long strings of binary inside a computer.

Introduced 3-5 · Unit 1

code

Code is an agreed mapping that tells the computer what a string of ones and zeros means. Without shared codes like ASCII for letters or RGB for colors, binary would just be meaningless switches.

3-56-89-12
Age-specific explanations
3-5

Code is an agreed mapping that tells the computer what a string of ones and zeros means. Without shared codes like ASCII for letters or RGB for colors, binary would just be meaningless switches.

6-8

Code is an agreed mapping that tells the computer what a string of ones and zeros means. Without shared codes like ASCII for letters or RGB for colors, binary would just be meaningless switches.

9-12

Code is an agreed mapping that tells the computer what a string of ones and zeros means. Without shared codes like ASCII for letters or RGB for colors, binary would just be meaningless switches.

Introduced 3-5 · Unit 1

input

Input is data going into a computer or program. Pressing a key, tapping a screen, uploading a photo — all input. A smart machine starts with input before it can process or decide anything.

3-56-89-12
Age-specific explanations
3-5

Input is data going into a computer or program. Pressing a key, tapping a screen, uploading a photo — all input. A smart machine starts with input before it can process or decide anything.

6-8

Input is data going into a computer or program. Pressing a key, tapping a screen, uploading a photo — all input. A smart machine starts with input before it can process or decide anything.

9-12

Input is data going into a computer or program. Pressing a key, tapping a screen, uploading a photo — all input. A smart machine starts with input before it can process or decide anything.

Introduced 3-5 · Unit 1

output

Output is what a computer sends back out — text on a screen, sound from a speaker, a printed page. Output is the result of processing input.

3-56-89-12
Age-specific explanations
3-5

Output is what a computer sends back out — text on a screen, sound from a speaker, a printed page. Output is the result of processing input.

6-8

Output is what a computer sends back out — text on a screen, sound from a speaker, a printed page. Output is the result of processing input.

9-12

Output is what a computer sends back out — text on a screen, sound from a speaker, a printed page. Output is the result of processing input.

Introduced 3-5 · Unit 1

processor

The processor is the part of a computer that does the work on data — math, logic, lookups. Input goes in, the processor runs instructions, and output comes out.

3-56-89-12
Age-specific explanations
3-5

The processor is the part of a computer that does the work on data — math, logic, lookups. Input goes in, the processor runs instructions, and output comes out.

6-8

The processor is the part of a computer that does the work on data — math, logic, lookups. Input goes in, the processor runs instructions, and output comes out.

9-12

The processor is the part of a computer that does the work on data — math, logic, lookups. Input goes in, the processor runs instructions, and output comes out.

Introduced 3-5 · Unit 1

RGB

RGB stands for red, green, and blue — the three color channels screens mix to make every color. Each channel is a number. Change the numbers, change the pixel.

3-56-89-12
Age-specific explanations
3-5

RGB stands for red, green, and blue — the three color channels screens mix to make every color. Each channel is a number. Change the numbers, change the pixel.

6-8

RGB stands for red, green, and blue — the three color channels screens mix to make every color. Each channel is a number. Change the numbers, change the pixel.

9-12

RGB stands for red, green, and blue — the three color channels screens mix to make every color. Each channel is a number. Change the numbers, change the pixel.

Introduced 3-5 · Unit 1

switch

A switch has two states: on or off. Computers contain billions of tiny switches. Everything digital — text, photos, games — is built from switches flipping on and off at incredible speed.

3-56-89-12
Age-specific explanations
3-5

A switch has two states: on or off. Computers contain billions of tiny switches. Everything digital — text, photos, games — is built from switches flipping on and off at incredible speed.

6-8

A switch has two states: on or off. Computers contain billions of tiny switches. Everything digital — text, photos, games — is built from switches flipping on and off at incredible speed.

9-12

A switch has two states: on or off. Computers contain billions of tiny switches. Everything digital — text, photos, games — is built from switches flipping on and off at incredible speed.

Introduced 3-5 · Unit 2

algorithm

An algorithm is a clear, ordered set of steps to finish a task. A recipe is an algorithm. So is a sorting procedure. Computers follow algorithms exactly — which is why each step must be precise.

3-56-89-12
Age-specific explanations
3-5

An algorithm is a clear, ordered set of steps to finish a task. A recipe is an algorithm. So is a sorting procedure. Computers follow algorithms exactly — which is why each step must be precise.

6-8

An algorithm is a clear, ordered set of steps to finish a task. A recipe is an algorithm. So is a sorting procedure. Computers follow algorithms exactly — which is why each step must be precise.

9-12

An algorithm is a clear, ordered set of steps to finish a task. A recipe is an algorithm. So is a sorting procedure. Computers follow algorithms exactly — which is why each step must be precise.

Introduced 3-5 · Unit 2

conditional

A conditional is an if-then choice. If the stack is empty, then stop. If the badge is valid, then unlock. Conditionals let programs branch instead of doing the same thing every time.

3-56-89-12
Age-specific explanations
3-5

A conditional is an if-then choice. If the stack is empty, then stop. If the badge is valid, then unlock. Conditionals let programs branch instead of doing the same thing every time.

6-8

A conditional is an if-then choice. If the stack is empty, then stop. If the badge is valid, then unlock. Conditionals let programs branch instead of doing the same thing every time.

9-12

A conditional is an if-then choice. If the stack is empty, then stop. If the badge is valid, then unlock. Conditionals let programs branch instead of doing the same thing every time.

Introduced 3-5 · Unit 2

debug

Debug means find the broken step, fix it, and test again. Programs rarely work on the first try. Debugging is not failure — it is how reliable software gets built.

3-56-89-12
Age-specific explanations
3-5

Debug means find the broken step, fix it, and test again. Programs rarely work on the first try. Debugging is not failure — it is how reliable software gets built.

6-8

Debug means find the broken step, fix it, and test again. Programs rarely work on the first try. Debugging is not failure — it is how reliable software gets built.

9-12

Debug means find the broken step, fix it, and test again. Programs rarely work on the first try. Debugging is not failure — it is how reliable software gets built.

Introduced 3-5 · Unit 2

iterative design

Iterative design is build, test, learn, improve — repeat. Debug a robot's steps. Retrain a model on better data. Ship a feature, measure harm, fix it. One pass is never enough.

3-56-89-12
Age-specific explanations
3-5

Iterative design is build, test, learn, improve — repeat. Debug a robot's steps. Retrain a model on better data. Ship a feature, measure harm, fix it. One pass is never enough.

6-8

Iterative design is build, test, learn, improve — repeat. Debug a robot's steps. Retrain a model on better data. Ship a feature, measure harm, fix it. One pass is never enough.

9-12

Iterative design is build, test, learn, improve — repeat. Debug a robot's steps. Retrain a model on better data. Ship a feature, measure harm, fix it. One pass is never enough.

Introduced 3-5 · Unit 2

loop

A loop is a step that repeats. Instead of writing stack a cup five times, you write repeat five times. Loops save space and match how real programs handle large jobs.

3-56-89-12
Age-specific explanations
3-5

A loop is a step that repeats. Instead of writing stack a cup five times, you write repeat five times. Loops save space and match how real programs handle large jobs.

6-8

A loop is a step that repeats. Instead of writing stack a cup five times, you write repeat five times. Loops save space and match how real programs handle large jobs.

9-12

A loop is a step that repeats. Instead of writing stack a cup five times, you write repeat five times. Loops save space and match how real programs handle large jobs.

Introduced 3-5 · Unit 2

sequence

Sequence means the order steps run in. Pour before you grab a cup and you make a mess. Change the sequence, change the result. Order matters in every algorithm.

3-56-89-12
Age-specific explanations
3-5

Sequence means the order steps run in. Pour before you grab a cup and you make a mess. Change the sequence, change the result. Order matters in every algorithm.

6-8

Sequence means the order steps run in. Pour before you grab a cup and you make a mess. Change the sequence, change the result. Order matters in every algorithm.

9-12

Sequence means the order steps run in. Pour before you grab a cup and you make a mess. Change the sequence, change the result. Order matters in every algorithm.

Introduced 3-5 · Unit 3

data

Data is collected information — numbers, words, images, locations, messages. Patterns hide inside data. Machine learning systems are only as useful as the data people give them.

3-56-89-12
Age-specific explanations
3-5

Data is collected information — numbers, words, images, locations, messages. Patterns hide inside data. Machine learning systems are only as useful as the data people give them.

6-8

Data is collected information — numbers, words, images, locations, messages. Patterns hide inside data. Machine learning systems are only as useful as the data people give them.

9-12

Data is collected information — numbers, words, images, locations, messages. Patterns hide inside data. Machine learning systems are only as useful as the data people give them.

Introduced 3-5 · Unit 3

diverse

Diverse data includes many different kinds of cases — people, objects, situations, edge cases. Diversity in training data helps a model work beyond the most common examples it saw.

3-56-89-12
Age-specific explanations
3-5

Diverse data includes many different kinds of cases — people, objects, situations, edge cases. Diversity in training data helps a model work beyond the most common examples it saw.

6-8

Diverse data includes many different kinds of cases — people, objects, situations, edge cases. Diversity in training data helps a model work beyond the most common examples it saw.

9-12

Diverse data includes many different kinds of cases — people, objects, situations, edge cases. Diversity in training data helps a model work beyond the most common examples it saw.

Introduced 3-5 · Unit 3

machine learning

Machine learning is when a computer finds patterns in examples instead of following rules a person wrote by hand. You provide training data; the system builds a model. That model makes predictions on new inputs.

3-56-89-12
Age-specific explanations
3-5

Machine learning is when a computer learns a pattern from examples instead of following steps a person typed.

6-8

Machine learning is when a computer finds patterns in examples instead of following rules a person wrote by hand. You provide training data; the system builds a model. That model makes predictions on new inputs.

9-12

Machine learning is function approximation from data — parameters learned to minimize loss on training data while generalizing beyond it.

Introduced 3-5 · Unit 3

prediction

A prediction is a guess about something you have not observed yet, based on a pattern in data you have. Predictions can be wrong. Good systems track how often they are right.

3-56-89-12
Age-specific explanations
3-5

A prediction is a guess about something you have not observed yet, based on a pattern in data you have. Predictions can be wrong. Good systems track how often they are right.

6-8

A prediction is a guess about something you have not observed yet, based on a pattern in data you have. Predictions can be wrong. Good systems track how often they are right.

9-12

A prediction is a guess about something you have not observed yet, based on a pattern in data you have. Predictions can be wrong. Good systems track how often they are right.

Introduced 3-5 · Unit 3

sorting

Sorting means arranging items into order — by size, date, name, or category. Sorting reveals structure in data. Many algorithms and machine learning steps depend on organized data.

3-56-8
Age-specific explanations
3-5

Sorting means arranging items into order — by size, date, name, or category. Sorting reveals structure in data. Many algorithms and machine learning steps depend on organized data.

6-8

Sorting means arranging items into order — by size, date, name, or category. Sorting reveals structure in data. Many algorithms and machine learning steps depend on organized data.

Introduced 3-5 · Unit 3

training data

Training data is the set of examples a model learns from — images with labels, rows in a spreadsheet, text pairs. The model inherits the strengths and blind spots of that data.

3-56-89-12
Age-specific explanations
3-5

Training data is the set of examples a model learns from — images with labels, rows in a spreadsheet, text pairs. The model inherits the strengths and blind spots of that data.

6-8

Training data is the set of examples a model learns from — images with labels, rows in a spreadsheet, text pairs. The model inherits the strengths and blind spots of that data.

9-12

Training data is the set of examples a model learns from — images with labels, rows in a spreadsheet, text pairs. The model inherits the strengths and blind spots of that data.

Introduced 3-5 · Unit 4

bias

Bias, in data, is a gap — some people or cases were left out or underrepresented. The model learns that gap faithfully. Bias is usually fixable by improving the data, not by blaming the machine.

3-56-89-12
Age-specific explanations
3-5

Bias is a gap in the data. It left some people or things out. Better data fixes it.

6-8

Bias, in data, is a gap — some people or cases were left out or underrepresented. The model learns that gap faithfully. Bias is usually fixable by improving the data, not by blaming the machine.

9-12

Bias spans coverage gaps, skewed labels, and disparate error rates at deployment. Measure per group; fix curation, not the symptom alone.

Introduced 3-5 · Unit 4

represent

To represent means the data stands in for the real world you care about. If your dataset has only one type of shoe, it does not represent all shoes. Poor representation leads to poor results.

3-56-89-12
Age-specific explanations
3-5

To represent means the data stands in for the real world you care about. If your dataset has only one type of shoe, it does not represent all shoes. Poor representation leads to poor results.

6-8

To represent means the data stands in for the real world you care about. If your dataset has only one type of shoe, it does not represent all shoes. Poor representation leads to poor results.

9-12

To represent means the data stands in for the real world you care about. If your dataset has only one type of shoe, it does not represent all shoes. Poor representation leads to poor results.

Introduced 3-5 · Unit 4

trustworthy

Trustworthy data is reliable because you know where it came from, how it was collected, and whether it was checked. Do not train a model on data you cannot explain.

3-56-89-12
Age-specific explanations
3-5

Trustworthy data is reliable because you know where it came from, how it was collected, and whether it was checked. Do not train a model on data you cannot explain.

6-8

Trustworthy data is reliable because you know where it came from, how it was collected, and whether it was checked. Do not train a model on data you cannot explain.

9-12

Trustworthy data is reliable because you know where it came from, how it was collected, and whether it was checked. Do not train a model on data you cannot explain.

6-8

Introduced in 6-8

Introduced 6-8 · Unit 1

act

To act is to execute an output — show a label, unlock a door, call a tool, send a message. Smart systems close the loop: sense, reason, act.

K-26-89-12
Age-specific explanations
K-2

To act is to execute an output — show a label, unlock a door, call a tool, send a message. Smart systems close the loop: sense, reason, act.

6-8

To act is to execute an output — show a label, unlock a door, call a tool, send a message. Smart systems close the loop: sense, reason, act.

9-12

To act is to execute an output — show a label, unlock a door, call a tool, send a message. Smart systems close the loop: sense, reason, act.

Introduced 6-8 · Unit 1

AND

AND means both conditions must be true for the result to be true. Valid badge AND door closed — both required to unlock. If either is false, the output is false.

6-89-12
Age-specific explanations
6-8

AND means both conditions must be true for the result to be true. Valid badge AND door closed — both required to unlock. If either is false, the output is false.

9-12

AND means both conditions must be true for the result to be true. Valid badge AND door closed — both required to unlock. If either is false, the output is false.

Introduced 6-8 · Unit 1

Boolean operator

Boolean operators — AND, OR, NOT — combine true and false values to make a decision. AND needs both true. OR needs at least one. NOT flips the value. They appear in hardware gates and in code.

6-89-12
Age-specific explanations
6-8

Boolean operators — AND, OR, NOT — combine true and false values to make a decision. AND needs both true. OR needs at least one. NOT flips the value. They appear in hardware gates and in code.

9-12

Boolean operators — AND, OR, NOT — combine true and false values to make a decision. AND needs both true. OR needs at least one. NOT flips the value. They appear in hardware gates and in code.

Introduced 6-8 · Unit 1

chip

A chip is a small piece of silicon packed with transistors and circuits. Your phone, laptop, and the servers running large AI models all depend on chips doing math with switches.

6-89-12
Age-specific explanations
6-8

A chip is a small piece of silicon packed with transistors and circuits. Your phone, laptop, and the servers running large AI models all depend on chips doing math with switches.

9-12

A chip is a small piece of silicon packed with transistors and circuits. Your phone, laptop, and the servers running large AI models all depend on chips doing math with switches.

Introduced 6-8 · Unit 1

hardware

Hardware is the physical parts — chips, memory, screens, sensors. If you can touch it or it draws power, it is hardware.

6-89-12
Age-specific explanations
6-8

Hardware is the physical parts — chips, memory, screens, sensors. If you can touch it or it draws power, it is hardware.

9-12

Hardware is the physical parts — chips, memory, screens, sensors. If you can touch it or it draws power, it is hardware.

Introduced 6-8 · Unit 1

inclusive OR

Inclusive OR is true when either input is true, including when both are true. Do not treat OR as exclusive — both-on is still a yes.

6-89-12
Age-specific explanations
6-8

Inclusive OR is true when either input is true, including when both are true. Do not treat OR as exclusive — both-on is still a yes.

9-12

Inclusive OR is true when either input is true, including when both are true. Do not treat OR as exclusive — both-on is still a yes.

Introduced 6-8 · Unit 1

logic gate

A logic gate is a small circuit that combines on-off signals with a rule — like AND, OR, or NOT. Gates are how hardware implements the same true-false decisions you write in code.

6-89-12
Age-specific explanations
6-8

A logic gate is a small circuit that combines on-off signals with a rule — like AND, OR, or NOT. Gates are how hardware implements the same true-false decisions you write in code.

9-12

A logic gate is a small circuit that combines on-off signals with a rule — like AND, OR, or NOT. Gates are how hardware implements the same true-false decisions you write in code.

Introduced 6-8 · Unit 1

NOT

NOT flips a value. True becomes false. False becomes true. It is the simplest gate — one input, one inverted output.

6-89-12
Age-specific explanations
6-8

NOT flips a value. True becomes false. False becomes true. It is the simplest gate — one input, one inverted output.

9-12

NOT flips a value. True becomes false. False becomes true. It is the simplest gate — one input, one inverted output.

Introduced 6-8 · Unit 1

OR

OR means at least one condition is true. Inclusive OR is true if one is true or if both are true. Fire alarm OR motion sensor — either can trigger the alert.

6-89-12
Age-specific explanations
6-8

OR means at least one condition is true. Inclusive OR is true if one is true or if both are true. Fire alarm OR motion sensor — either can trigger the alert.

9-12

OR means at least one condition is true. Inclusive OR is true if one is true or if both are true. Fire alarm OR motion sensor — either can trigger the alert.

Introduced 6-8 · Unit 1

reason

To reason, in AI terms, is to process input with rules or a learned model to reach a conclusion. It is not human reasoning — it is computation on representations — but it fills the middle step between sense and act.

6-89-12
Age-specific explanations
6-8

To reason, in AI terms, is to process input with rules or a learned model to reach a conclusion. It is not human reasoning — it is computation on representations — but it fills the middle step between sense and act.

9-12

To reason, in AI terms, is to process input with rules or a learned model to reach a conclusion. It is not human reasoning — it is computation on representations — but it fills the middle step between sense and act.

Introduced 6-8 · Unit 1

sense

To sense is to gather input from the environment — the machine equivalent of noticing. Cameras, microphones, and sensors sense. No sensing, no smart behavior.

K-26-89-12
Age-specific explanations
K-2

To sense is to gather input from the environment — the machine equivalent of noticing. Cameras, microphones, and sensors sense. No sensing, no smart behavior.

6-8

To sense is to gather input from the environment — the machine equivalent of noticing. Cameras, microphones, and sensors sense. No sensing, no smart behavior.

9-12

To sense is to gather input from the environment — the machine equivalent of noticing. Cameras, microphones, and sensors sense. No sensing, no smart behavior.

Introduced 6-8 · Unit 1

silicon

Silicon is the material microchips are made from. It can act as a semiconductor — sometimes conducting electricity, sometimes blocking it. That on-off behavior is the physical basis of computing.

6-89-12
Age-specific explanations
6-8

Silicon is the material microchips are made from. It can act as a semiconductor — sometimes conducting electricity, sometimes blocking it. That on-off behavior is the physical basis of computing.

9-12

Silicon is the material microchips are made from. It can act as a semiconductor — sometimes conducting electricity, sometimes blocking it. That on-off behavior is the physical basis of computing.

Introduced 6-8 · Unit 1

software

Software is the instructions that tell hardware what to do — apps, operating systems, AI models loaded into memory. A frozen app is usually software, not a broken chip.

6-89-12
Age-specific explanations
6-8

Software is the instructions that tell hardware what to do — apps, operating systems, AI models loaded into memory. A frozen app is usually software, not a broken chip.

9-12

Software is the instructions that tell hardware what to do — apps, operating systems, AI models loaded into memory. A frozen app is usually software, not a broken chip.

Introduced 6-8 · Unit 1

transistor

A transistor is a tiny electronic switch on a chip. Billions of them turn on and off to run programs. Modern AI runs on hardware built from transistors switching at extreme speed.

6-89-12
Age-specific explanations
6-8

A transistor is a tiny electronic switch on a chip. Billions of them turn on and off to run programs. Modern AI runs on hardware built from transistors switching at extreme speed.

9-12

A transistor is a tiny electronic switch on a chip. Billions of them turn on and off to run programs. Modern AI runs on hardware built from transistors switching at extreme speed.

Introduced 6-8 · Unit 1

troubleshoot

Troubleshoot means narrow down where a problem lives — connection, software, settings, hardware — and test one layer at a time. Random guessing wastes time; systematic checks fix things faster.

6-89-12
Age-specific explanations
6-8

Troubleshoot means narrow down where a problem lives — connection, software, settings, hardware — and test one layer at a time. Random guessing wastes time; systematic checks fix things faster.

9-12

Troubleshoot means narrow down where a problem lives — connection, software, settings, hardware — and test one layer at a time. Random guessing wastes time; systematic checks fix things faster.

Introduced 6-8 · Unit 1

truth table

A truth table lists every combination of inputs and the output a gate or condition produces. It is the cheat sheet for AND, OR, and NOT — every row is a test case.

6-89-12
Age-specific explanations
6-8

A truth table lists every combination of inputs and the output a gate or condition produces. It is the cheat sheet for AND, OR, and NOT — every row is a test case.

9-12

A truth table lists every combination of inputs and the output a gate or condition produces. It is the cheat sheet for AND, OR, and NOT — every row is a test case.

Introduced 6-8 · Unit 2

algorithmic bias

Algorithmic bias is when a model is reliably more wrong for some group than others. Nobody typed in prejudice — the model learned skew from skewed data. The fix starts with data, labels, and evaluation across groups.

6-89-12
Age-specific explanations
6-8

Algorithmic bias is when a model is reliably more wrong for some group than others. Nobody typed in prejudice — the model learned skew from skewed data. The fix starts with data, labels, and evaluation across groups.

9-12

Algorithmic bias is when a model is reliably more wrong for some group than others. Nobody typed in prejudice — the model learned skew from skewed data. The fix starts with data, labels, and evaluation across groups.

Introduced 6-8 · Unit 2

artificial intelligence

Artificial intelligence is software that senses input, finds patterns or plans actions, and outputs decisions — from photo sorting to chatbots to agents. Today's AI is mostly pattern learning at scale, not human-like understanding.

6-89-12
Age-specific explanations
6-8

Artificial intelligence is software that senses input, finds patterns or plans actions, and outputs decisions — from photo sorting to chatbots to agents. Today's AI is mostly pattern learning at scale, not human-like understanding.

9-12

Artificial intelligence is software that senses input, finds patterns or plans actions, and outputs decisions — from photo sorting to chatbots to agents. Today's AI is mostly pattern learning at scale, not human-like understanding.

Introduced 6-8 · Unit 2

feature

A feature is a measurable property of an example the model can use — sweetness, color, word count, pixel brightness. Models do not see objects the way you do; they see lists of features.

6-89-12
Age-specific explanations
6-8

A feature is a measurable property of an example the model can use — sweetness, color, word count, pixel brightness. Models do not see objects the way you do; they see lists of features.

9-12

A feature is a measurable property of an example the model can use — sweetness, color, word count, pixel brightness. Models do not see objects the way you do; they see lists of features.

Introduced 6-8 · Unit 2

generalization

Generalization is the ability to perform well beyond the training set. It is the main goal of machine learning. Without generalization, you just have a lookup table.

6-89-12
Age-specific explanations
6-8

Generalization is the ability to perform well beyond the training set. It is the main goal of machine learning. Without generalization, you just have a lookup table.

9-12

Generalization is the ability to perform well beyond the training set. It is the main goal of machine learning. Without generalization, you just have a lookup table.

Introduced 6-8 · Unit 2

generalize

To generalize means the model works on new data it never trained on — not just memorizing old examples. A useful model generalizes. One that only succeeds on training data failed to learn the real pattern.

6-89-12
Age-specific explanations
6-8

To generalize means the model works on new data it never trained on — not just memorizing old examples. A useful model generalizes. One that only succeeds on training data failed to learn the real pattern.

9-12

To generalize means the model works on new data it never trained on — not just memorizing old examples. A useful model generalizes. One that only succeeds on training data failed to learn the real pattern.

Introduced 6-8 · Unit 2

label

A label is the correct answer attached to an example — cat or dog, spam or not spam. Labeled data tells the model what to predict. Bad labels teach bad habits.

6-89-12
Age-specific explanations
6-8

A label is the correct answer attached to an example — cat or dog, spam or not spam. Labeled data tells the model what to predict. Bad labels teach bad habits.

9-12

A label is the correct answer attached to an example — cat or dog, spam or not spam. Labeled data tells the model what to predict. Bad labels teach bad habits.

Introduced 6-8 · Unit 2

labeled dataset

A labeled dataset is a collection of examples where each item has its answer attached. Supervised learning runs on labeled datasets. The quality of labels is as important as the quantity.

6-89-12
Age-specific explanations
6-8

A labeled dataset is a collection of examples where each item has its answer attached. Supervised learning runs on labeled datasets. The quality of labels is as important as the quantity.

9-12

A labeled dataset is a collection of examples where each item has its answer attached. Supervised learning runs on labeled datasets. The quality of labels is as important as the quantity.

Introduced 6-8 · Unit 2

model

A model is the pattern a system learned from data and saved for reuse. It takes input, applies what it learned, and outputs a guess. Different data builds a different model.

6-89-12
Age-specific explanations
6-8

A model is the pattern a system learned from data and saved for reuse. It takes input, applies what it learned, and outputs a guess. Different data builds a different model.

9-12

A model is the pattern a system learned from data and saved for reuse. It takes input, applies what it learned, and outputs a guess. Different data builds a different model.

Introduced 6-8 · Unit 2

overfitting

Overfitting is when a model memorizes training examples instead of learning the underlying pattern. It scores high on old data and fails on new data. Too little diversity or too much complexity often causes it.

6-89-12
Age-specific explanations
6-8

Overfitting is when a model memorizes training examples instead of learning the underlying pattern. It scores high on old data and fails on new data. Too little diversity or too much complexity often causes it.

9-12

Overfitting is when a model memorizes training examples instead of learning the underlying pattern. It scores high on old data and fails on new data. Too little diversity or too much complexity often causes it.

Introduced 6-8 · Unit 3

activation function

An activation function decides how a neuron responds after the weighted sum — a smooth curve instead of a hard on-off cutoff. It lets networks learn non-linear patterns. Without it, deep stacks would collapse to simple math.

9-12
Age-specific explanations
6-8

An activation function decides how a neuron responds after the weighted sum — a smooth curve instead of a hard on-off cutoff. It lets networks learn non-linear patterns. Without it, deep stacks would collapse to simple math.

9-12

An activation function decides how a neuron responds after the weighted sum — a smooth curve instead of a hard on-off cutoff. It lets networks learn non-linear patterns. Without it, deep stacks would collapse to simple math.

Introduced 6-8 · Unit 3

backpropagation

Backpropagation sends error backward through the network to calculate how much each weight contributed. Then gradient descent updates those weights. It is the standard way deep networks learn from mistakes.

9-12
Age-specific explanations
6-8

Backpropagation sends error backward through the network to calculate how much each weight contributed. Then gradient descent updates those weights. It is the standard way deep networks learn from mistakes.

9-12

Backpropagation sends error backward through the network to calculate how much each weight contributed. Then gradient descent updates those weights. It is the standard way deep networks learn from mistakes.

Introduced 6-8 · Unit 3

deep learning

Deep learning uses neural networks with many layers to learn complex patterns — images, speech, language. More layers can capture more detail, but they need more data and more compute.

6-89-12
Age-specific explanations
6-8

Deep learning uses neural networks with many layers to learn complex patterns — images, speech, language. More layers can capture more detail, but they need more data and more compute.

9-12

Deep learning uses neural networks with many layers to learn complex patterns — images, speech, language. More layers can capture more detail, but they need more data and more compute.

Introduced 6-8 · Unit 3

gradient descent

Gradient descent is how training finds better weights automatically. It measures error, figures out which direction reduces error, and nudges weights that way — step after step. It is hill-climbing in reverse on a loss landscape.

9-12
Age-specific explanations
6-8

Gradient descent is how training finds better weights automatically. It measures error, figures out which direction reduces error, and nudges weights that way — step after step. It is hill-climbing in reverse on a loss landscape.

9-12

Gradient descent is how training finds better weights automatically. It measures error, figures out which direction reduces error, and nudges weights that way — step after step. It is hill-climbing in reverse on a loss landscape.

Introduced 6-8 · Unit 3

layer

A layer is a row of neurons whose outputs feed the next row. Stacking layers lets a network build complex patterns from simple decisions. Deep learning means many layers.

6-89-12
Age-specific explanations
6-8

A layer is a row of neurons whose outputs feed the next row. Stacking layers lets a network build complex patterns from simple decisions. Deep learning means many layers.

9-12

A layer is a row of neurons whose outputs feed the next row. Stacking layers lets a network build complex patterns from simple decisions. Deep learning means many layers.

Introduced 6-8 · Unit 3

neural network

A neural network is layers of connected neurons that learn by tuning weights. Input flows forward, the network produces an output, and training adjusts weights to reduce mistakes. It is pattern matching at scale.

6-89-12
Age-specific explanations
6-8

A neural network is layers of neurons that learn by tuning weights. Weighted sum, threshold, repeat — stacked thousands of times.

9-12

A neural network is a differentiable stack of matrix operations trained via backpropagation and gradient descent.

Introduced 6-8 · Unit 3

neuron

In a neural network, a neuron is a tiny math unit — multiply inputs by weights, add them up, compare to a threshold, output on or off. It is arithmetic, not a brain cell. The name is historical.

6-89-12
Age-specific explanations
6-8

In a neural network, a neuron is a tiny math unit — multiply inputs by weights, add them up, compare to a threshold, output on or off. It is arithmetic, not a brain cell. The name is historical.

9-12

In a neural network, a neuron is a tiny math unit — multiply inputs by weights, add them up, compare to a threshold, output on or off. It is arithmetic, not a brain cell. The name is historical.

Introduced 6-8 · Unit 3

sum

The sum is what you get after multiplying each input by its weight and adding everything together. That total gets compared to a threshold to decide if the neuron fires.

6-89-12
Age-specific explanations
6-8

The sum is what you get after multiplying each input by its weight and adding everything together. That total gets compared to a threshold to decide if the neuron fires.

9-12

The sum is what you get after multiplying each input by its weight and adding everything together. That total gets compared to a threshold to decide if the neuron fires.

Introduced 6-8 · Unit 3

threshold

A threshold is the cutoff the weighted sum must beat for a neuron to fire. Below the threshold, output off. At or above, output on. Later models replace hard thresholds with smoother activation functions.

6-89-12
Age-specific explanations
6-8

A threshold is the cutoff the weighted sum must beat for a neuron to fire. Below the threshold, output off. At or above, output on. Later models replace hard thresholds with smoother activation functions.

9-12

A threshold is the cutoff the weighted sum must beat for a neuron to fire. Below the threshold, output off. At or above, output on. Later models replace hard thresholds with smoother activation functions.

Introduced 6-8 · Unit 3

weight

A weight is a number saying how much an input matters to a neuron. Bigger weight, stronger pull on the output. Training a network mostly means adjusting weights until errors shrink.

6-89-12
Age-specific explanations
6-8

A weight is a number saying how much an input matters to a neuron. Bigger weight, stronger pull on the output. Training a network mostly means adjusting weights until errors shrink.

9-12

A weight is a number saying how much an input matters to a neuron. Bigger weight, stronger pull on the output. Training a network mostly means adjusting weights until errors shrink.

Introduced 6-8 · Unit 4

accuracy

Accuracy is the fraction of predictions that were correct — right divided by total. Ten right out of sixteen is about sixty-three percent. Accuracy alone does not tell you who the errors hurt.

6-89-12
Age-specific explanations
6-8

Accuracy is the fraction of predictions that were correct — right divided by total. Ten right out of sixteen is about sixty-three percent. Accuracy alone does not tell you who the errors hurt.

9-12

Accuracy is the fraction of predictions that were correct — right divided by total. Ten right out of sixteen is about sixty-three percent. Accuracy alone does not tell you who the errors hurt.

Introduced 6-8 · Unit 4

clean data

Clean data has duplicates removed, typos fixed, blanks handled, and labels corrected. Dirty data poisons training and makes every metric look untrustworthy. Clean before you trust a score.

6-89-12
Age-specific explanations
6-8

Clean data has duplicates removed, typos fixed, blanks handled, and labels corrected. Dirty data poisons training and makes every metric look untrustworthy. Clean before you trust a score.

9-12

Clean data has duplicates removed, typos fixed, blanks handled, and labels corrected. Dirty data poisons training and makes every metric look untrustworthy. Clean before you trust a score.

Introduced 6-8 · Unit 4

curate

To curate data is to choose what goes in, label it, and remove junk. Curation is a human decision. What you include, exclude, and how you label shapes the model's behavior.

6-89-12
Age-specific explanations
6-8

To curate data is to choose what goes in, label it, and remove junk. Curation is a human decision. What you include, exclude, and how you label shapes the model's behavior.

9-12

To curate data is to choose what goes in, label it, and remove junk. Curation is a human decision. What you include, exclude, and how you label shapes the model's behavior.

Introduced 6-8 · Unit 4

data quality

Data quality covers accuracy, completeness, consistency, and relevance. High-quality data is clean, labeled well, and covers the cases that matter. Models inherit data quality — they cannot fix bad source material alone.

6-89-12
Age-specific explanations
6-8

Data quality covers accuracy, completeness, consistency, and relevance. High-quality data is clean, labeled well, and covers the cases that matter. Models inherit data quality — they cannot fix bad source material alone.

9-12

Data quality covers accuracy, completeness, consistency, and relevance. High-quality data is clean, labeled well, and covers the cases that matter. Models inherit data quality — they cannot fix bad source material alone.

Introduced 6-8 · Unit 4

error

An error is a case the model got wrong. Errors are not random noise — they cluster. Always ask which examples fail and whether failures fall harder on one group.

6-89-12
Age-specific explanations
6-8

An error is a case the model got wrong. Errors are not random noise — they cluster. Always ask which examples fail and whether failures fall harder on one group.

9-12

An error is a case the model got wrong. Errors are not random noise — they cluster. Always ask which examples fail and whether failures fall harder on one group.

Introduced 6-8 · Unit 4

evaluation

Evaluation means measuring how well a model performs on data it was not trained on. You pick metrics, run tests, and inspect errors. Building a model without evaluation is guessing.

6-89-12
Age-specific explanations
6-8

Evaluation means measuring how well a model performs on data it was not trained on. You pick metrics, run tests, and inspect errors. Building a model without evaluation is guessing.

9-12

Evaluation means measuring how well a model performs on data it was not trained on. You pick metrics, run tests, and inspect errors. Building a model without evaluation is guessing.

Introduced 6-8 · Unit 4

representativeness

Representativeness asks whether your data looks like the real situations the model will face. Training on one campus and deploying everywhere fails if the data is not representative. Measure gaps before you deploy.

6-89-12
Age-specific explanations
6-8

Representativeness asks whether your data looks like the real situations the model will face. Training on one campus and deploying everywhere fails if the data is not representative. Measure gaps before you deploy.

9-12

Representativeness asks whether your data looks like the real situations the model will face. Training on one campus and deploying everywhere fails if the data is not representative. Measure gaps before you deploy.

Introduced 6-8 · Unit 5

digital citizenship

Digital citizenship is using technology responsibly — protecting privacy, spotting scams, giving credit, and thinking about how your actions affect others online. It is ethics plus security plus everyday habits.

6-89-12
Age-specific explanations
6-8

Digital citizenship is using technology responsibly — protecting privacy, spotting scams, giving credit, and thinking about how your actions affect others online. It is ethics plus security plus everyday habits.

9-12

Digital citizenship is using technology responsibly — protecting privacy, spotting scams, giving credit, and thinking about how your actions affect others online. It is ethics plus security plus everyday habits.

Introduced 6-8 · Unit 5

encryption

Encryption scrambles data so only someone with the key can read it. Even if a message is intercepted, ciphertext looks like garbage without decryption. It protects privacy in transit and at rest.

6-89-12
Age-specific explanations
6-8

Encryption scrambles data so only someone with the key can read it. Even if a message is intercepted, ciphertext looks like garbage without decryption. It protects privacy in transit and at rest.

9-12

Encryption scrambles data so only someone with the key can read it. Even if a message is intercepted, ciphertext looks like garbage without decryption. It protects privacy in transit and at rest.

Introduced 6-8 · Unit 5

malware

Malware is harmful software that steals data, spies, or locks your files. It often arrives through bad downloads or phishing links. Updates and antivirus add layers — no single tool is enough.

6-89-12
Age-specific explanations
6-8

Malware is harmful software that steals data, spies, or locks your files. It often arrives through bad downloads or phishing links. Updates and antivirus add layers — no single tool is enough.

9-12

Malware is harmful software that steals data, spies, or locks your files. It often arrives through bad downloads or phishing links. Updates and antivirus add layers — no single tool is enough.

Introduced 6-8 · Unit 5

phishing

Phishing is a fake message designed to steal passwords or personal info — urgent emails, bogus login pages, scam texts. Real companies rarely ask you to verify a password by link. When in doubt, do not click.

6-89-12
Age-specific explanations
6-8

Phishing is a fake message designed to steal passwords or personal info — urgent emails, bogus login pages, scam texts. Real companies rarely ask you to verify a password by link. When in doubt, do not click.

9-12

Phishing is a fake message designed to steal passwords or personal info — urgent emails, bogus login pages, scam texts. Real companies rarely ask you to verify a password by link. When in doubt, do not click.

Introduced 6-8 · Unit 5

privacy

Privacy is control over who sees and uses your data. Photos, location, messages, browsing history — all personal. Before you share, ask who gets access and whether you can take it back.

6-89-12
Age-specific explanations
6-8

Privacy is control over who sees and uses your data. Photos, location, messages, browsing history — all personal. Before you share, ask who gets access and whether you can take it back.

9-12

Privacy is control over who sees and uses your data. Photos, location, messages, browsing history — all personal. Before you share, ask who gets access and whether you can take it back.

Introduced 6-8 · Unit 5

risk vs. benefit

Risk versus benefit means weighing what could go wrong against what you gain before you act. Posting a photo might get likes but expose location. Good digital citizens do this calculation on purpose.

6-89-12
Age-specific explanations
6-8

Risk versus benefit means weighing what could go wrong against what you gain before you act. Posting a photo might get likes but expose location. Good digital citizens do this calculation on purpose.

9-12

Risk versus benefit means weighing what could go wrong against what you gain before you act. Posting a photo might get likes but expose location. Good digital citizens do this calculation on purpose.

Introduced 6-8 · Unit 5

security

Security is protecting systems and data from unauthorized access, theft, and damage. Strong passwords, updates, encryption, and careful sharing are layers. Security is never one switch — it is a stack.

6-89-12
Age-specific explanations
6-8

Security is protecting systems and data from unauthorized access, theft, and damage. Strong passwords, updates, encryption, and careful sharing are layers. Security is never one switch — it is a stack.

9-12

Security is protecting systems and data from unauthorized access, theft, and damage. Strong passwords, updates, encryption, and careful sharing are layers. Security is never one switch — it is a stack.

9-12

Introduced in 9-12

Introduced 9-12 · Unit 1

embeddings

Embeddings are vectors that capture meaning so similar things sit close together in math space. King and queen end up near each other because their contexts overlap. Language models start by turning tokens into embeddings.

9-12

Introduced 9-12 · Unit 1

representation

Representation is how real-world things become numbers a computer can process. Letters become ASCII. Colors become RGB. Words and meanings later become vectors and embeddings. AI runs on representations, not on raw reality.

3-56-89-12
Age-specific explanations
3-5

Representation is how real-world things become numbers a computer can process. Letters become ASCII. Colors become RGB. Words and meanings later become vectors and embeddings. AI runs on representations, not on raw reality.

6-8

Representation is how real-world things become numbers a computer can process. Letters become ASCII. Colors become RGB. Words and meanings later become vectors and embeddings. AI runs on representations, not on raw reality.

9-12

Representation is how real-world things become numbers a computer can process. Letters become ASCII. Colors become RGB. Words and meanings later become vectors and embeddings. AI runs on representations, not on raw reality.

Introduced 9-12 · Unit 1

vectors

A vector is an ordered list of numbers representing something — a point in space, a set of features, a word's meaning. Models compare vectors with math — distance, angle, weighted sums — to make predictions.

9-12

Introduced 9-12 · Unit 2

loss

Loss is a number measuring how wrong the model's predictions are during training. Training tries to push loss down. Watch the loss curve — if training loss drops but test loss rises, you are overfitting.

9-12

Introduced 9-12 · Unit 3

attention

Attention lets a model focus on the parts of input that matter most for the current prediction. When translating or answering, not every word weighs equally. Attention assigns those weights dynamically.

9-12

Introduced 9-12 · Unit 3

transformers

Transformers are a neural network architecture built around attention. They process sequences — especially text — by letting each piece look at relevant context in parallel. Most modern language models are transformers.

9-12

Introduced 9-12 · Unit 4

emergent behavior

Emergent behavior is a capability that appears at large scale though it was not explicitly programmed — reasoning steps, tool use patterns, unexpected fluency. It is real but unpredictable. Treat emergent skills as test, not trust.

9-12

Introduced 9-12 · Unit 4

inference

Inference is running a trained model on new input to produce a prediction. Training happens once; inference happens every time a user asks. Inference cost and speed matter in real products.

9-12

Introduced 9-12 · Unit 4

large language model

A large language model is a transformer trained on massive text to predict the next token. Scale brings broad language skill — writing, coding, summarizing — but also hallucination risk and no built-in truth checking.

9-12

Introduced 9-12 · Unit 4

pretraining

Pretraining is the first training phase on huge general datasets — books, web text, code. The model learns language structure before anyone fine-tunes it for a specific task. Pretraining is expensive and done once.

9-12

Introduced 9-12 · Unit 4

probability

Probability is a number from zero to one describing how likely something is. Models output probabilities, not certainties. Treat high-confidence wrong answers as a reminder — statistics, not truth.

9-12

Introduced 9-12 · Unit 4

tokens

Tokens are the chunks a language model reads and writes — whole words, parts of words, or punctuation. Text becomes a token sequence. Models have token limits; long documents must be split or summarized.

9-12

Introduced 9-12 · Unit 5

context

Context is everything the model can see in the current window — your prompt, prior messages, retrieved documents. Models have no memory beyond that window unless you build memory outside. Context quality drives answer quality.

9-12

Introduced 9-12 · Unit 5

prompting

Prompting is how you instruct a language model — clear task, examples, format, constraints. Good prompts reduce garbage outputs. Prompting is not magic; it is structured communication with a statistical system.

9-12

Introduced 9-12 · Unit 6

caching

Caching stores earlier computation so you do not recompute it. In AI systems, caching repeated prompts or key-value states cuts cost and latency. Tradeoff: stale cache can serve outdated answers if data changed.

9-12

Introduced 9-12 · Unit 6

cost

Cost in AI systems is mostly compute — GPU time, tokens processed, storage, API calls. Longer prompts, bigger models, and more users multiply cost. Engineering means measuring cost per task, not ignoring it.

9-12

Introduced 9-12 · Unit 6

latency

Latency is delay — time from request to response. Big models, long contexts, and slow networks increase latency. Product design balances smarter models against how long users will wait.

9-12

Introduced 9-12 · Unit 7

agent

An agent is a system that perceives context, decides actions, and acts — often in a loop with tools. It is notice, reason, act at software scale. Agents need guardrails because wrong actions have real consequences.

K-26-89-12
Age-specific explanations
K-2

A smart machine that notices and decides is the kid version of what older students call an agent.

6-8

An agent senses, reasons, and acts — the grown-up name for a smart machine.

9-12

An agent loops: perceive state, plan, call tools, observe results, repeat. Guardrails and human oversight belong in the design.

Introduced 9-12 · Unit 7

perceive

Perceive means taking in raw signals and turning them into structured input the system can use. Perception is sense plus processing — pixels become features, sound becomes tokens.

9-12

Introduced 9-12 · Unit 7

planning

Planning is breaking a goal into steps before executing. Agents plan which tools to call in what order. Plans can fail — good systems re-plan when a step returns an error instead of blindly continuing.

9-12

Introduced 9-12 · Unit 7

tools

Tools, for AI agents, are external actions a model can call — search, calculator, database lookup, send email. The model proposes a tool call; your code runs it and returns results. Tools connect language to the real world.

9-12

Introduced 9-12 · Unit 8

grounding

Grounding ties a model's answer to verified sources — retrieved docs, databases, citations. Ungrounded models invent plausible text. Grounding reduces hallucination by anchoring outputs to evidence you supply.

9-12

Introduced 9-12 · Unit 8

retrieval

Retrieval fetches relevant documents or data at query time instead of relying only on what the model memorized in training. Search over a knowledge base, pull top matches, inject them into context. That keeps answers current.

9-12

Introduced 9-12 · Unit 9

accountability

Accountability assigns responsibility when something goes wrong. Blaming the algorithm ends the conversation too early. Ask who chose the data, who deployed it, and who monitors it now.

9-12

Introduced 9-12 · Unit 9

algorithmic accountability

Algorithmic accountability means people who build and deploy systems answer for outcomes — audits, documentation, appeal paths, monitoring after launch. The model computes; humans remain responsible for harm it causes.

9-12

Introduced 9-12 · Unit 9

fairness metrics

Fairness metrics compare model performance across groups — error rates, precision, recall by demographic or region. One overall accuracy can hide harm to a minority. Disaggregate before you declare success.

9-12

Introduced 9-12 · Unit 9

harm

Harm is real damage from a system — wrongful denial, privacy breach, misinformation, discrimination. Risk analysis asks who gets hurt and how badly. Low average performance can still cause serious harm to a few.

6-89-12
Age-specific explanations
6-8

Harm is real damage from a system — wrongful denial, privacy breach, misinformation, discrimination. Risk analysis asks who gets hurt and how badly. Low average performance can still cause serious harm to a few.

9-12

Harm is real damage from a system — wrongful denial, privacy breach, misinformation, discrimination. Risk analysis asks who gets hurt and how badly. Low average performance can still cause serious harm to a few.

Introduced 9-12 · Unit 9

harm analysis

Harm analysis systematically asks what could go wrong, who bears the risk, and how severe failures would be. Run it before shipping an AI feature — especially when decisions affect grades, hiring, or safety.

9-12

Introduced 9-12 · Unit 9

precision

Precision measures how many of the model's positive calls were actually correct. High precision means when it says yes, it is usually right — but it might miss many yes cases. Use it alongside recall.

9-12

Introduced 9-12 · Unit 9

recall

Recall measures how many of the actual positives the model caught. High recall finds most cases — but may flag many false alarms. Fair evaluation reports both precision and recall, often split by group.

9-12

Introduced 9-12 · Unit 10

academic integrity

Academic integrity is doing your own honest work according to the rules — no cheating, no hidden AI, no copied answers. Policies vary by assignment. When unsure, ask before you submit.

3-59-12
Age-specific explanations
3-5

Academic integrity is doing your own honest work according to the rules — no cheating, no hidden AI, no copied answers. Policies vary by assignment. When unsure, ask before you submit.

9-12

Academic integrity is doing your own honest work according to the rules — no cheating, no hidden AI, no copied answers. Policies vary by assignment. When unsure, ask before you submit.

Introduced 9-12 · Unit 10

attribution

Attribution means naming your sources and helpers — quotes, images, AI tools. It builds trust and avoids plagiarism. Say what the machine did and what you did.

K-29-12
Age-specific explanations
K-2

Attribution means naming your sources and helpers — quotes, images, AI tools. It builds trust and avoids plagiarism. Say what the machine did and what you did.

9-12

Attribution means naming your sources and helpers — quotes, images, AI tools. It builds trust and avoids plagiarism. Say what the machine did and what you did.

Introduced 9-12 · Unit 10

authorship

Authorship is who created a work and gets credit. If AI drafted it, you still owe transparency about your role — editing, fact-checking, ideas. Claiming full authorship for machine output is dishonest.

K-29-12
Age-specific explanations
K-2

Authorship is who created a work and gets credit. If AI drafted it, you still owe transparency about your role — editing, fact-checking, ideas. Claiming full authorship for machine output is dishonest.

9-12

Authorship is who created a work and gets credit. If AI drafted it, you still owe transparency about your role — editing, fact-checking, ideas. Claiming full authorship for machine output is dishonest.

Introduced 9-12 · Unit 10

intellectual property

Intellectual property covers legal rights to creations — copyright on text and art, patents on inventions, trademarks on brands. AI training on copyrighted work and AI-generated output raise open legal questions. Know your school's policy.

9-12

Introduced 9-12 · Unit 11

AI careers

AI careers span more than coding — data curation, evaluation, policy, UX, hardware, education, ethics review. Core skills stay valuable: clear thinking, statistics, communication, and knowing what machines cannot do alone.

9-12

Introduced 9-12 · Unit 11

alignment

Alignment is the effort to make AI systems pursue goals humans actually want — helpful, honest, harmless — not just maximize a training score. Misaligned systems optimize the metric and break the intent. Alignment is an ongoing research and policy problem.

9-12

Introduced 9-12 · Unit 11

data governance

Data governance is the rules for collecting, storing, sharing, and deleting data — who can access it, how long it is kept, what consent was given. Good governance prevents leaks and builds trust.

9-12

Introduced 9-12 · Unit 11

policy

Policy is the rules governments, schools, and companies set for AI use — privacy law, classroom bans, disclosure requirements, safety testing. Technology moves fast; policy catches up. Know the rules where you live and learn.

9-12

Introduced 9-12 · Unit 11

security engineering

Security engineering designs systems that stay safe under attack — encryption, access control, logging, patching, least privilege. It is proactive building, not panic after a breach.

9-12

Introduced 9-12 · Unit 11

threat model

A threat model lists what you are protecting, who might attack it, and how. Build defenses against real threats, not every hypothetical. Schools might prioritize phishing and data leaks over nation-state hacking.

9-12