Patterns, not magic.
A model learns patterns from examples and uses them to produce an output. A convincing answer still needs checking.
Explore the explanation ↗AI Juicing / The channel
Less watching the future.
More making stuff with it.
Come along for the experiments, the useful discoveries, and the bits that don’t quite go to plan. Start with the basics, then put them to work.
Loading behind the camera…
Pick a path. Press play.
A little less mystery.
A lot more “I can do that.”
AI Basics
Get comfortable with the ideas behind AI, try running it locally, and learn how to give it useful context and instructions.
Lesson 1 of 4 · 4:20
Understand how AI works using a simple dot.
An approachable introduction to models, patterns, and next-word prediction, using points on a graph instead of heavy maths.
Give it a go
Watch on YouTube or join the comments ↗Explain what an AI model does in one sentence. Include why a confident answer can still be wrong.
Before the next rabbit hole
Three ideas to take with you.
Give it a useful brief.
Say what you need, share the context, and describe the result you want. Review the first attempt and refine your instructions.
Explore prompt writing ↗Bring the source along.
Retrieval-augmented generation finds relevant material and gives it to the model as context. It helps ground an answer; it doesn’t guarantee accuracy.
Explore RAG ↗From watching to making
Curious what happens next?
Take a look at the tools and experiments growing out of all this tinkering.