AI & LLM foundations
What AI, machine learning, and large language models actually are, how they work at a topline level, and where their capabilities end.
Every executive now has an opinion about AI. Most of those opinions are built on hype, headlines, and a demo someone saw once. This block fixes that. You will finally understand what an AI actually is, how a large language model works under the hood, and why that matters for every bet you are about to place with your budget and your reputation.
We start with the mechanics. What is AI, what is machine learning, and what is a large language model, explained without the academic fog. Then we walk through what happens the moment you hit enter: tokens, training, and inference. You will learn why the context window is the model's working memory, and why that single concept explains half the frustration your teams have with these tools.
From there we get honest about power and limits. You will see what LLMs are genuinely great at, and where they fall flat on their face. We tackle hallucinations head on, because a confident wrong answer is far more dangerous than an obvious one, and your organization is already acting on those answers. Finally, we strip away the myth of machine understanding and show you what a model truly does: prediction, not comprehension.
Why does a senior leader need this? Because you cannot govern, fund, or challenge what you do not understand. When you know how the machine thinks, you ask sharper questions, you spot vendor nonsense in seconds, and you set policy that survives contact with reality. This is not a coding course. It is the mental model that lets you lead the AI conversation instead of nodding along in it.
What you'll master
- Explain the difference between AI, machine learning, and large language models in plain business language
- Trace what happens from prompt to answer, including tokens, training, and inference
- Use the context window concept to diagnose why AI tools succeed or fail on your tasks
- Judge where an LLM adds real value and where it becomes a liability
- Detect and mitigate hallucinations before they reach a decision or a customer
- Challenge vendor claims by understanding that models predict rather than understand
- Set informed expectations for teams deploying AI across the business
Modules
Covers the core mechanics of how AI, machine learning, and LLMs process and produce text.
Explains what LLMs do well, where they fail, and why confident answers can still be wrong.
Frequently asked questions
What does the AI & LLM foundations block cover?
AI & LLM foundations covers what AI, machine learning and large language models actually are, what happens between a prompt and an answer, and where these systems break down. It is organised in two modules of three lessons each: one on the mechanics (tokens, training, inference, context window), one on capabilities, limits and hallucinations.
Do I need a technical background to follow it?
No. AI & LLM foundations is written for senior leaders who fund, govern or challenge AI projects, not for engineers. There is no code and no maths: the goal is a mental model solid enough to ask sharper questions and spot vendor nonsense.
What is the difference between AI, machine learning and a large language model?
AI is the broad ambition of getting machines to perform tasks we associate with intelligence; machine learning is one way to get there, by having systems learn patterns from data rather than following hand-written rules; a large language model is a specific machine learning system trained to predict text. Being able to state that distinction in plain business language is the first outcome of the AI & LLM foundations block.
Why does the context window matter to a manager rather than to a developer?
The context window is the model's working memory: everything it can hold in view while producing an answer. It explains a large share of the frustration teams have with AI tools, because once a document, a conversation or a dataset exceeds that window, the model quietly loses part of it. Knowing this lets you diagnose failures instead of blaming the tool or the team.
How do you deal with hallucinations before they reach a customer?
You treat a confident answer as a claim to verify, not a result to publish. AI & LLM foundations devotes a full lesson to hallucinations, because a plausible wrong answer is far more dangerous than an obviously wrong one, and covers how to detect them and where to place human checks before an answer feeds a decision or reaches a customer.
Where does this block sit in the AI Essentials track, and should I start with it?
AI & LLM foundations is the foundational block of the AI Essentials track, and it is the right starting point if you have opinions about AI but no clear picture of how a model produces an answer. Everything that follows, from prompting to agentic AI, assumes you know that a model predicts rather than understands.