AI Foundations ← pitcsolutions.com All lessons

AI Foundations

From “what is AI?” to how ChatGPT writes its answers. 13 short, illustrated lessons with hands-on demos and quick quizzes, written for students starting from zero.

By Pushpjeet Cholkar. About 2.6 hours in total, at your own pace.

Start lesson 1 See all lessons

Your course map

Course map The 13 lessons drawn as a network of connected neurons, grouped into six tracks from left to right. Select a node to open that lesson. Start here Machine learning Deep learning Transformers and LLMs Generative models AI on AWS 1AI family 2Learning types 3Algorithms 4Train vs infer 5Deep learning 6Vision 7Language 8Transformers 9LLM intuition 10LLM internals 11GANs 12VAEs 13AWS services

Each circle is a lesson, connected like neurons in a network. Follow the bold path in order, or jump to any topic. Lessons you've opened turn green.

Start here

What AI is and the different ways machines learn.

  1. 1AI vs ML vs Deep Learning vs Generative AIHow the four terms fit inside each other, with one banking example.10 min
  2. 2Types of Machine LearningSupervised, unsupervised, semi-supervised and reinforcement learning.12 min

Machine learning

Choosing algorithms, and the difference between learning and using.

  1. 3Machine Learning AlgorithmsWhich algorithm to use for which problem, plus the interview favourites.15 min
  2. 4Training vs InferenceLearning from data versus using what was learned.8 min

Deep learning

Neural networks, and how they let computers see and read.

  1. 5Deep Learning FundamentalsNeurons, activation, backpropagation and the tricks that make deep nets work.20 min
  2. 6Computer VisionHow computers turn pixels into understanding.12 min
  3. 7Natural Language ProcessingHow computers read, understand and generate human language.12 min

Transformers and LLMs

The ideas behind ChatGPT, Claude and Gemini.

  1. 8Transformers and AttentionThe architecture behind ChatGPT, Claude and Gemini.15 min
  2. 9How LLMs Work: The IntuitionA story-first walkthrough: patterns, next-token prediction and tokens.12 min
  3. 10How LLMs Work: Under the HoodEmbeddings, self-attention, generation and the three stages of training.15 min

Generative models

Networks that create brand-new images, faces and more.

  1. 11Generative Adversarial Networks (GANs)Two networks compete: a faker and a judge.8 min
  2. 12Variational Autoencoders (VAEs)Compress data into a smooth map, then generate from it.10 min

AI on AWS

Mapping every idea in this course to real cloud services.

  1. 13AWS Services for AI, ML and GenAIWhich AWS service belongs to which category, and which to learn first.10 min

How each lesson works

Read

Plain-language explanations with diagrams, analogies and real examples. No maths degree needed.

Try

Small interactive demos: roll a ball down a loss curve, watch attention, generate text token by token.

Discuss

Classroom activities you can do alone or with friends, with suggested answers to check yourself.

Check

A short quiz at the end of every lesson with instant feedback and explanations.

Your reading progress is saved in this browser only.