What We Actually Cover in an AI Basics Team Training

Before we hand a team any AI tool, we teach them how the thing actually works. Not because it's academically interesting — because a team that understands the basics asks better questions, spots the right use cases faster, and stops expecting AI to be either magic or a search engine. Here's what's actually in our AI Fundamentals session.

Most AI training that lands in front of Australian small and medium businesses skips straight to "here are ten prompts to try." We start a step earlier, because the businesses that get the most out of AI aren't the ones with the fanciest tools — they're the ones whose people understand what they're working with.

What "AI" actually means

We open by pulling apart a term that gets used to mean almost anything. In plain terms: AI is software that performs tasks normally requiring human judgement, improves from examples rather than fixed rules, and generates a prediction or response — it doesn't "know" facts the way a person does. That last point alone reframes how a room thinks about the tools they're about to use.

Untangling AI, machine learning and deep learning

These three get used interchangeably in almost every boardroom conversation, and they're not the same thing. We walk through them as nested circles: AI is the broad goal, machine learning is a subset of AI where systems learn from data instead of being explicitly programmed, and deep learning is a subset of machine learning using layered neural networks — the technology behind today's large language models, image generators and voice AI.

How a model actually "learns," and why tokens matter

We cover the five-step loop behind every model in use today — collect data, train, find patterns, test, predict and improve — and then get concrete about tokens: the small chunks of text a model actually reads and writes, roughly three-quarters of a word each. This isn't trivia. Tokens drive what a business pays, how fast a tool responds, and how much a model can "see" at once — understanding that changes how a team writes prompts and sets expectations about cost and limits.

The model landscape, and who's actually using what

We walk teams through the major model families and what each is genuinely best at — Claude for careful reasoning and long documents, GPT as a broad general-purpose assistant, Gemini for its Google Workspace integration, Copilot for its embedding inside Word, Excel, Outlook and Teams, Llama for businesses that want to run models on their own infrastructure, plus the specialists in image and voice generation. We back this with current adoption data for the US and Australia, so the session isn't abstract — it's grounded in what other businesses are actually using right now, including where Australia's numbers diverge from the US (Copilot's enterprise reach, for one).

Data and privacy, covered honestly

Every session includes a plain-language walkthrough of how each major model handles the data you put into it — what's used for training by default, what isn't, and what changes on a paid business tier versus a free consumer account. Teams leave with a concrete short list: use business or API tiers, turn off "improve the model" toggles on personal accounts used for work, and keep client data and financial records out of consumer AI tools.

Bringing it back to the job

The session closes by tying all of it back to how people actually work day to day: prompting more effectively once you understand how a model "reads" text, setting realistic expectations because AI predicts rather than guarantees facts, and spotting the tasks — repetitive, pattern-based, language-heavy — where AI genuinely helps. That's deliberately where the foundations session ends and not where the engagement ends — it's the groundwork for a hands-on follow-up session built around each team's actual workflows.

The takeaway

Skipping straight to tools without this grounding is how businesses end up with a handful of enthusiasts using AI well and everyone else either ignoring it or misusing it. An hour spent on fundamentals first is what makes every session after it land.