I was tired of buzzword-heavy AI talks, and marginally impactful projects. Surely we can do some more inspiring things with these LLMs!
Physical AI is where all those breakthroughs finally reach into the real world: robots that see, move, and figure things out for themselves. I think it's the most exciting frontier in tech right now. It's also genuinely hard to break into, because it isn't one field but five stacked on top of each other: electronics, mechanics, programming, data, and AI.

Eight months ago I started my own robotics journey from scratch, and I was completely overwhelmed. How do you get from an LED blink to humanoids that do your dishes? There's thousands of scattered tutorials with no sense of what came first, or what any of it was building toward. So I decided to make the map.

Inspired by my favorite engineering and strategy games (Satisfactory, Planet Crafter, Civ Six), I built a tech tree for robotics: a structured, visual path that takes you from pure curiosity all the way toward advanced humanoids, one unlocked skill at a time. Those games are proof that we'll happily spend hours mastering a complex system when it's laid out as a satisfying series of unlocks. So why not point that same instinct at learning something real?

In this talk I'll share the guide I wish I'd had. How I took an intimidating, multidisciplinary field and structured it into a dependency graph. How I consolidated nearly everything I've built across my career into one learning path. And how the community has already started building it with me. We're not reinventing all this knowledge - we're simply taking amazing resources and making them all more accessible. This is an open-source community project to help navigate the chaos.

You'll leave with a mental model for making any complex field approachable, and hopefully the itch to start your own robotics journey.