Series · 22 of 22 parts · Beginner

The Groundwork

Start from absolute zero. The computer science, machine learning, and neural network ideas AI is built on, explained in plain words with no equations and no numbers, so every AI term makes sense before you meet it.

What you'll understand

By the end you can watch any AI explainer and follow every word: tokens, embeddings, loss, gradient descent, and what a model really is.

Episodes

01What a Computer Actually IsBefore any AI, a computer is just a fast, patient helper that follows instructions exactly. Start here, from nothing.4 min 02On, Off, and Everything BetweenUnderneath everything, a computer holds just one tiny thing: a switch that is on or off. Meet the bit.5 min 03Turning the World Into SwitchesHow a photo, a song, and your words all become patterns of on and off that a machine can hold.5 min 04A Recipe the Machine FollowsA program is a recipe for a very literal cook. Meet the algorithm, with no code in sight.6 min 05What We Mean by a PatternSpotting patterns is the heart of everything AI does. What a pattern really is, in plain words.5 min 06Turning Meaning Into PlacesMachines handle meaning by turning things into places, so similar things sit close together. Meet embeddings.5 min 07Guessing What Comes NextA huge part of intelligence is just guessing what comes next. The one trick behind modern AI.5 min 08How Likely, ReallyLikelihood is just how much something would surprise you. Probability and temperature, gently.6 min 09Two Ways to Make a Machine SmartWrite every rule yourself, or show the machine examples and let it learn. What machine learning means.4 min 10Learning From ExamplesGuess, get corrected, repeat across a mountain of examples. How training actually works.5 min 11Getting It Wrong on PurposeMistakes are the teacher. How being wrong is the engine that makes a machine learn.4 min 12Practice Makes It SmallerSmall steps downhill toward fewer mistakes. Gradient descent, with a beautiful video to watch.4 min 13The Knobs InsideLearning is just turning a vast bank of tiny knobs until the machine behaves. Meet parameters.4 min 14Loosely Inspired by the BrainNeural networks borrowed one small idea from the brain. What that comparison does and doesn't mean.4 min 15Tiny Decision Makers in LayersMany tiny parts, each making one small decision, arranged in layers. Meet neurons and layers.5 min 16Why Deeper Is DifferentStack many layers deep and ideas build on ideas. Where the phrase deep learning comes from.4 min 17Studying vs Taking the TestA model has two lives: slow, costly learning, then fast, cheap answering. Training versus inference.5 min 18What We Mean by a ModelA model is the finished, trained thing. What people really mean by the model, and by language model.4 min 19Breaking Language Into PiecesBefore a model reads text, it chops it into pieces. Meet the token and the tokenizer.4 min 20Words as Places in MeaningEvery piece of language gets a spot on a map of meaning. Embeddings, now named for words.5 min 21Careful or AdventurousShould a model play it safe or take a risk on its next word? The temperature dial, explained.5 min 22You're Ready. What Comes NextA warm send off. Everything the main AI series leans on now has a home in your head.5 min

Opening the library

The Groundwork

What you'll understand

Episodes

Keep up with AI. One email a week.

The Groundwork

What you'll understand

Episodes

Keep up with AI. One email a week.