Week Two - LLM Basic Flashcards
What does “LLM” stand for, and what is it?
Large Language Model – An AI trained on massive amounts of text to generate and understand human-like language.
How does an LLM actually work?
It predicts the next word in a sentence based on everything it’s seen before. It’s not thinking—it’s pattern prediction.
What kind of data are LLMs trained on?
Books, websites, code, articles, forums—billions of words from across the internet and curated sources.
What makes LLMs different from traditional AI systems?
LLMs are open-ended—they can write, chat, summarize, or explain. Old AI was task-specific (like spam filters or chess).
What’s a “token” in an LLM and why does it matter?
A token is a word—or part of a word. Models have token limits, which control how long their responses can be.
What’s an example of a tool powered by an LLM?
ChatGPT is an LLM. But so are tools like Jasper, Notion AI, or Grammarly—just with different interfaces and purposes.
What are common use cases for LLMs?
Writing, research, code generation, chatbots, sales outreach, summarizing, tutoring, product ideas, customer support.
Can LLMs actually understand what you’re saying?
No—they don’t truly understand. They mirror understanding by recognizing word patterns based on training.
What are key risks or limitations of LLMs?
They can “hallucinate” (make stuff up), repeat bias, and give wrong answers that sound confident.
What’s the difference between an LLM and an AI tool that uses one?
The LLM is the engine. Tools (like a chatbot or writing app) use that engine via API to power different experiences.
