Explain how context and likely next tokens shape an LLM response.
Week 01 · Foundations
AI-Assisted Creation Course map + what is an LLM?
Start with one practical question: how can an AI produce useful language while still needing people to verify its claims and control its actions?
Before class
Nothing is required. You do not need to watch a video, create an account, install software, or complete a reading before Week 1.
By the end
Four ideas you can explain and use.
Week 1 builds a working mental model. You are not expected to memorize technical vocabulary.
Separate confident language from claims supported by evidence.
Describe how goals, tools, and a loop turn an LLM into an agent system.
Identify actions that should stop for human review or approval.
Class route
Two classes, each 1 hour 20 minutes.
Each meeting has its own start, learning checkpoint, and stopping point. Times describe the current classroom plan.
Meeting 1
Build the LLM mental model.
Welcome + course map
See the semester route and Week 1 expectations.
Task map + email spine
Connect AI literacy to one practical task.
Context without oversharing
Build the minimum sufficient fictional context.
LLM model
Prediction, truth, tokens, and tools.
Visible context + video
Check what the product can use, then watch the short excerpt.
Evidence-card exit
Write the two-sentence mental model and close.
Meeting 2
Verify outputs and preview agents.
Reconnect
Rebuild the Meeting 1 model together.
Training + assistant behavior
Compare pretraining, post-training, base models, and assistants.
Claim/evidence lab
Classify four claims and correct their strength.
System + permission decisions
Label components, choose authority, and decide whether to proceed, ask, or block.
Evidence-card exit
Revise the mental model and choose an optional weekend assignment route.
Class materials
Open the lesson—or jump to the part you need.
The presentation remains the complete classroom experience. These links simply make it easier to return to a topic.
Primary class material · 22 teaching screens · 12 + 10
Week 1 interactive lesson
Follow the instructor-led sequence for demonstrations, quick checks, a verification activity, and a short preview of agent systems.
Watch or replay
See next-token prediction instead of only hearing about it.
The short video reinforces the same concepts taught in class. The privacy-enhanced player loads only after you choose it.
After class
Optional weekend assignment: choose Quick or Extended.
If you want additional practice this weekend, complete one of the two assignment routes below. Week 2 begins with an in-class recap, and eCampus will identify the concepts included on Quiz 1.
No after-class work
Choose this route with no penalty. Keep your in-class evidence card only if it helps your Week 2 recap.
Quick Review
Watch the short 3Blue1Brown visual explanation and write one question you still have.
Extended Review
Watch the selected Karpathy segments and answer the six guided prompts. Allow additional time for writing.
Deep Dive
Watch the complete Karpathy lecture and use all 18 checkpoints. This is beyond the optional weekend assignment.
Go deeper
References stay available without crowding the class route.
Use these only when you want more background or a more technical explanation.