AI Fundamentals: Your Guide to Getting Started

Most AI confusion isn't conceptual - it's vocabulary. Here are the terms you'll actually encounter as a beginner, explained without the jargon.

Model

The trained system that produces an output from your input - text, image, or video. Different models are trained differently and produce different results from the same prompt.

Prompt

The instruction you give a model - written text, an uploaded image, or both. A more specific prompt generally produces a more predictable result than a vague one.

Generation

One output from one prompt. Most workflows involve generating several times and picking or refining the best result, not getting it perfect on the first attempt.

Token

The small chunks of text a language model actually processes - roughly a word or part of a word. It's mostly relevant when you hit a length limit on a prompt or output.

Fine-tuning

Training an existing model further on a specific, narrower dataset so it performs better on a specific task or style, rather than training a new model from scratch.

Inference

The act of actually running a trained model to produce a result - as opposed to "training," which is the earlier, much more expensive process of building the model in the first place.

You don't need all of these to start creating - Promptus's interface is built so you can generate a first result with just a plain-language prompt, and pick up the terminology as it becomes relevant.