Generative AI and evaluating AI output

Generative AI creates new content, such as text, images, audio, or computer code, by learning patterns from enormous amounts of existing data. Large language models, which power many AI chat tools, generate text by repeatedly predicting a likely next piece of text based on everything before it. This makes their answers fluent, but not always correct: they can state false information confidently (often called hallucinations), reflect biases in their training data, give outdated answers, or produce code that looks right but has bugs or security flaws. Using AI responsibly means treating its output as a draft to check: verify facts against reliable sources, test and read any AI-generated code before using it, watch for bias, avoid pasting in private or personal information, and follow your school's or workplace's rules about when AI may be used and how to disclose it.

This explanation is AI-generated and can occasionally be wrong. 🚩 Report