Key Ideas

Here is a set of ideas you can use at an upcoming event or with your friends.

Prompt:

What you type to guide the AI. You tell it the goal, the steps, and the limits.

LLM:

An LLM is an AI model that works with text. Examples include GPT-4/5, Gemini, Llama, Mistral, and Claude. ChatGPT is an app built on an LLM.

Token:

A small piece of text. It can be a word, a space, or part of a word. LLMs use tokens to predict what comes next.

Context window:

How much information a model can hold at one time. After that, things get summarized. Think of it like short-term memory.

Memory:

If memory is off, the AI forgets chat to chat. When the limit is hit, it keeps the key points, so it can feel longer than it is.

Prompt engineering:

Methods that make prompts clear and repeatable so you get useful results.
Example: “You are a friendly local AI coach. Write a 60-word invite for Newmarket small businesses to try AI over coffee. Keep it plain and encouraging.”

Prompt engineering:

Using the context window well. Choose and condense what you send, keep high-quality info, cut the fluff, and define things clearly so the model stays focused.

Join us in person. Check the Event Page for the next Journey With AI meetup.