3.2 Large Language Models (LLMs)

A Large Language Model (LLM) is a generative AI trained on enormous amounts of text. By reading billions of sentences, it learns how words and ideas usually go together. When you give it a prompt, it works mainly by predicting the next most likely word, again and again, until it has built a full, coherent answer.

Think of typing on your phone: when you start "Selam, endemin" the keyboard suggests "adderk". An LLM does the same thing but on a massive scale, so it can write a whole paragraph, translate Amharic to English, or summarise a long article.