Artificial intelligence, or AI, is software designed to perform tasks that usually require human thinking, such as recognizing images, translating language, or making recommendations. Machine learning is one important way to build AI by letting a computer find patterns in data instead of programming every rule by hand. Understanding AI matters because it affects search engines, social media, health tools, cars, games, and schoolwork.
Separating myths from facts helps students use AI safely, fairly, and intelligently.
A machine learning system usually starts with data, then trains a model by adjusting numbers so its predictions get closer to the correct answers. Statistics and probability help the system measure uncertainty, compare patterns, and estimate how well it will work on new examples. AI does not truly understand the world like a person, and it can make mistakes when data are biased, incomplete, or different from what it saw during training.
Good AI design includes testing, human review, privacy protection, and clear limits on what the system should be trusted to do.
Key Facts
- AI is a broad field, while machine learning is a method within AI that learns patterns from data.
- A simple prediction model can be written as y = mx + b, where the model adjusts m and b to fit data.
- Model error can be measured with error = predicted value - actual value.
- Accuracy = correct predictions / total predictions.
- Probability values range from 0 to 1, where 0 means impossible and 1 means certain.
- Myth: AI is always objective. Fact: AI can reflect bias in its data, design, or use.
Vocabulary
- Artificial Intelligence
- Artificial intelligence is computer software designed to perform tasks that seem to require human reasoning, perception, or decision making.
- Machine Learning
- Machine learning is a way for computers to improve at a task by finding patterns in data.
- Training Data
- Training data are the examples used to teach a machine learning model how to make predictions or classifications.
- Model
- A model is a mathematical or computational rule that maps inputs to outputs based on learned patterns.
- Bias
- Bias is a systematic error that can cause an AI system to treat some cases or groups unfairly or inaccurately.
Common Mistakes to Avoid
- Thinking AI understands like a human. AI finds statistical patterns, so it can produce a confident answer without real understanding or common sense.
- Assuming more data always means better AI. More data helps only if the data are relevant, accurate, and representative of the problem.
- Judging a model only by training performance. A model can memorize training examples and still fail on new data, which is called overfitting.
- Treating AI answers as facts without checking. AI systems can make errors, repeat bias, or invent details, so important outputs should be verified with reliable sources.
Practice Questions
- 1 A spam filter tests 200 emails and correctly labels 170 of them. What is its accuracy?
- 2 An image model predicts that 45 out of 60 photos contain a dog, and 39 of those predictions are correct. What fraction of the dog predictions were correct, and what is the decimal value?
- 3 A face recognition system is trained mostly on photos taken in bright indoor lighting. Explain why it might perform poorly on outdoor night photos, and name one way engineers could improve it.