What Is AI?
What AI actually is
Artificial intelligence is the field of building computer systems that perform tasks we associate with human intelligence, such as recognising patterns, making predictions, or generating language. Most systems in use today are narrow AI: they are specialised for a defined task. General AI-a system that can transfer understanding across arbitrary tasks like a person-remains hypothetical. Do not let exam options imply that current chatbots are generally intelligent or conscious.
Machine learning (ML) is a subset of AI. Instead of hand-coding every rule, developers supply training data, and an algorithm adjusts a model so it captures statistical patterns. Inference is using that trained model on new input. Deep learning uses multi-layer neural networks; generative AI produces new text, images, or audio, often from patterns learned from large datasets.
Traps: an algorithm is a procedure, while a model is the learned artefact; training and inference are different stages; a system that only follows fixed if-then rules is automation, not necessarily ML; and a model finding a correlation does not prove causation. Also, “the AI decided” is shorthand-people set objectives, data, and thresholds.
Worked example. A spam filter is trained on emails labelled spam or not spam. It learns word and sender patterns, then scores a new message. It does not “know” spam; it estimates probability from past examples. Rule of thumb: for any AI claim, ask: what task, what training data, what output, and what human set the goal?