Artificial Intelligence: What It Is and How It Learns
Exploring AI, data, patterns, training, and responsible use
Artificial Intelligence: What It Is and How It Learns
Exploring AI, data, patterns, training, and responsible use
CS - Grade 6-8
- 1
In your own words, explain what artificial intelligence is. Include one example of an AI system you might see or use in everyday life.
Think about apps or devices that seem to make decisions or suggestions.
Artificial intelligence is technology that can perform tasks that usually require human thinking, such as recognizing speech, finding patterns, making predictions, or giving suggestions. One example is a music app that recommends songs based on what a person has listened to before. - 2
Classify each item as AI or not AI: a voice assistant that answers spoken questions, a regular calculator that adds two numbers, a video game character that changes strategy based on how you play, and a desk lamp with an on-off switch.
The voice assistant is AI because it understands speech and chooses an answer. The regular calculator is not AI because it follows a fixed rule for arithmetic. The video game character may be AI because it adapts its strategy based on the player. The desk lamp is not AI because it only turns on or off when controlled. - 3
A computer is trained to tell whether photos show cats or dogs. What kind of data would it need for training, and why would many examples be helpful?
Labeled data means examples that include the correct answer.
It would need many labeled photos of cats and dogs. Many examples are helpful because the computer can learn patterns, such as ear shape, face shape, fur, and body size, instead of memorizing just one picture. - 4
Explain the difference between a rule-based program and a machine learning system.
A rule-based program follows instructions that a human wrote step by step. A machine learning system learns patterns from data and uses those patterns to make predictions or decisions about new examples. - 5
Look at this pattern from a tiny training set: sunny days with high temperature were labeled ice cream sales high, and rainy days with low temperature were labeled ice cream sales low. Predict the likely sales for a sunny, hot day, and explain your reasoning.
Use the pattern in the examples, not outside information.
The likely sales would be high because the training examples show that sunny, hot weather is connected with high ice cream sales. The system would use that pattern to make its prediction. - 6
What is training data in machine learning?
Training data is the set of examples a machine learning system studies to find patterns. The examples may include inputs, such as images or words, and correct labels, such as cat, dog, spam, or not spam. - 7
A student says, "AI learns exactly like a human student learns." Explain why this statement is not completely correct.
Compare pattern-finding with human understanding.
The statement is not completely correct because AI does not understand experiences the same way humans do. Many AI systems learn by finding statistical patterns in data, while humans can use emotions, goals, common sense, and real-world experiences. - 8
An email filter marks some messages as spam. Name two features, or clues, the AI might use to decide whether an email is spam.
The AI might use clues such as suspicious links, repeated sales words, unknown senders, many exclamation points, or messages similar to past spam emails. These features help the system find patterns connected to spam. - 9
A school trains an AI homework helper using only essays written by high school students. What problem might happen if a sixth grader asks it for writing advice?
Think about whether the training examples match the user.
The AI might give advice that is too advanced or not appropriate for a sixth grader because the training data did not include enough writing from younger students. This shows that training data should match the people and tasks the AI will be used for. - 10
Explain what bias means in AI. Give one example of how biased training data could cause an unfair result.
Bias in AI means the system makes unfair or inaccurate decisions because the data or design favors some groups or patterns over others. For example, if a face recognition system is trained mostly on one skin tone, it may work worse for people with other skin tones. - 11
A model is tested after training. It correctly identifies 8 out of 10 animal pictures. What is its accuracy as a percent, and what does that mean?
Convert 8 out of 10 into a percent.
The accuracy is 80 percent because 8 out of 10 equals 80 out of 100. This means the model got 80 percent of the test pictures correct, but it still made mistakes on 20 percent of them. - 12
Why is it important to test an AI system with new examples that were not used during training?
It is important because testing with new examples shows whether the AI learned useful patterns or only memorized the training data. A good AI system should work well on new examples that are similar to the real situations it will face. - 13
A robot vacuum uses sensors to avoid walls and learn which areas of a room need cleaning most often. Describe one input, one output, and one possible decision the system makes.
Inputs are information coming in, and outputs are actions or information going out.
One input is sensor data about nearby walls or dirt. One output is the robot moving forward, turning, or changing speed. One possible decision is to spend more time cleaning an area where its sensors detect more dirt. - 14
An AI image generator creates a picture from a text prompt. Explain why the result may not be perfectly accurate.
The result may not be perfectly accurate because the AI creates an image based on patterns it learned from training data, not because it truly understands the world. It may mix up details, leave out important parts, or create objects that look possible but are wrong. - 15
Write two responsible-use rules a student should follow when using AI for schoolwork.
Think about honesty, safety, and checking information.
A student should use AI to support learning instead of copying answers and should check important information with reliable sources. A student should also follow the teacher's rules and be honest about when AI was used.