Artificial Intelligence Basics – Short Notes
Practical notes for choosing and using AI tools with careful human review.
What AI means
- Artificial intelligence describes computer systems designed to perform tasks that normally require human-like perception, language, prediction, or decision support.
- AI systems differ widely: a rule-based program, a classifier, and a generative model do not work in the same way.
- A model’s output depends on its design, training data, inputs, and operating context.
Training and use
- Machine-learning systems learn patterns from examples or feedback rather than relying only on hand-written rules.
- Training adjusts a model; inference is the use of a trained model to produce an output from new input.
- Some products combine models with search, databases, tools, or human review.
Strengths and limits
- AI can summarize, classify, predict, generate, and find patterns quickly when the task and data suit it.
- Outputs can be incomplete, biased, outdated, or confidently wrong; fluency is not proof of accuracy.
- Performance may vary across languages, populations, and unusual cases.
Use it thoughtfully
- Define the task and decide whether AI is appropriate before sharing data or accepting an output.
- Verify important claims using trustworthy evidence and a person who understands the subject.
- Keep human responsibility for decisions that affect people, safety, money, rights, or access.