Skip to enroll

AI/LLM/ML: The Definitive Guide

2 Learners
10 Lessons
0.0 (0)
6 hours
Advanced
Course

A clear map of artificial intelligence, machine learning, and large language models: how they learn, how transformers predict the next token, and how to judge the limits.

Overview

AI, machine learning, and large language models are three nested ideas, not three brands. AI is the broad goal of software that performs tasks we associate with judgment. Machine learning is one way to build that software, by fitting a model to examples. A large language model is a particular model that predicts the next token of text.

This guide is for people who want the mechanisms, not a product tour. You will walk from labels and loss to overfitting, then from tokens to transformers, and finally to the system around a model: evaluation, retrieval, tools, and risk.

You will not train a frontier model here. You will learn what each knob changes, which claims are testable, and which famous explanations to watch next. Two quizzes check the map before you use the vocabulary in a design review.

Features
  • A precise split between AI, machine learning, and large language models
  • A picture of features, labels, loss, and gradient updates
  • A test you can describe for overfit and underfit
  • A token-level view of language models and a one-diagram transformer
  • What fine-tuning changes, and what retrieval and tools add around the model
  • A study plan that separates marketing claims from mechanisms

Target audiences
  • Engineers and analysts who need the vocabulary to be exact
  • Technical leads reviewing AI proposals
  • Students who have used chat tools and want the machinery
  • People who will read papers or talks and need a map first

Requirements
  • Comfort with graphs and with the idea of a function that maps inputs to outputs
  • No calculus homework is collected, but you should be willing to follow a slope on a chart
  • No GPU and no paid API are required
  • The course is advanced in ideas, not in coding assignments

Curriculum

4 sections 12 lectures 6 hours

The map 2 lectures 38 mins
  • Three names, three jobs: AI, ML, and LLM
    Locked 18 min
  • Learning from examples, not rules alone
    Locked 20 min
How models learn 4 lectures 1h 18m
  • Features, labels, and a simple model
    Locked 22 min
  • Loss, gradients, and why training repeats
    Locked 24 min
  • Overfit, underfit, and a fair test
    Locked 20 min
  • Quiz: Model foundations
    Locked 12 min
Language models 3 lectures 1h 8m
  • Tokens, context, and next-word prediction
    Locked 22 min
  • Transformers in one diagram
    Locked 24 min
  • What fine-tuning changes, and what it does not
    Locked 22 min
Using them with judgment 3 lectures 54 mins
  • Evaluation, retrieval, and tool use
    Locked 22 min
  • Limits, risk, and a study plan
    Locked 20 min
  • Quiz: Guide review
    Locked 12 min

Frequently Asked Questions

01. Will I train a large model?

No. You will learn what training, fine-tuning, retrieval, and tool use each change.

02. Is this a math course?

It uses the ideas of loss and gradients in plain language and points you to visual chapters. It does not assign proofs.

03. Do I need Python?

No. Optional talks show code. The lessons stand without running it.

04. Are the diagrams from a textbook?

No. They are original sketches for this course. The videos are public talks by their own authors.

05. How long is the course?

About 6 hours, including 10 lessons and 2 quizzes.

Instructors

Liam Carter

Product Design Lead

  • 0.0 Instructor Rating
  • 2 Learners
  • 1 Course

I’m Liam Carter, a Product Design Lead passionate about creating digital products that are both useful and enjoyable to use. I enjoy turning complex ideas into simple, intuitive experiences through thoughtful design, research, and prototyping. In my courses, I share practical design methods, real-world workflows, and lessons I’ve learned from working with cross-functional teams. My goal is to help you develop the skills and confidence to design products that solve real problems.

Reviews

0.0

No ratings yet

5 0
4 0
3 0
2 0
1 0