From Zero to AI Agents
There is massive uncertainty in AI right now.

So what should you actually learn?
Warning
Tool fluency alone will not make you stand out.
Note
You do not need a PhD in AI. You do not need to build an LLM from scratch. That is Level 4 — a very different career path.
Most people in the industry don’t operate at Level 4. And most never need to.

Important
This course gives you fluency at Level 2 — a conceptual understanding of how AI actually works, so you can reason about it, evaluate it, and build with it, regardless of which tools survive.
Understand AI as a concept — not just manage the latest tool.
Before AI: Product → Design → Engineering was a waterfall. Three separate jobs, three separate teams.
With AI: the boundaries are dissolving. One person can now do what used to take a full team.

“I’m no leading expert, but I know something about AI and how to teach it.”

“Being a software developer is just being a glorified Google searcher.”
And I mean that. The barrier to understanding this material is much lower than you think.

Tip
You have all passed the GMAT. You have all come a long way to Berkeley. You have worked in very tough industries.
You are all more than clever enough to understand all of this.
I don’t expect you to memorize every concept. The goal is to eradicate unknown unknowns.
Unknown unknowns are dangerous — you can’t ask about what you don’t know exists. After this course, things will ring a bell. You’ll recognize concepts, know the right questions to ask, and know exactly where to go deeper when you need to.

Important
All slides from this course are public and always available. You can revisit any concept at any time.
“It’s better to know what you don’t know and be able to go back to it — than to not know what you don’t know.”
Build a map of what exists. You can always go deeper later.
Note
Next up: Class 1 — Introduction to Software
What is code, how does the web work, and how do you ship software to real users.
Four classes. Zero to AI agents. Let’s go.