Andrew “Boz” Bosworth's 10 rules for navigating the next design paradigm
Meta’s Chief Technology Officer on consciousness, questioning design paradigms, and building for the future.
Illustration by Luis Mazón
Andrew “Boz” Bosworth has been at Meta for almost 20 years, where he’s led some of the company’s most defining initiatives—from features like News Feed and Messenger to wearables like Meta Quest headsets, Ray-Ban Meta glasses, and Meta Orion glasses. As Meta’s current CTO and Head of Reality Labs, Boz is at the forefront of spatial computing and the next generation of human-computer interaction. At Config 2025, Figma’s annual conference for product builders, Boz sat down with Figma Co-Founder and CEO Dylan Field for a wide-ranging conversation. These 10 rules, distilled from their discussion, offer both a glimpse into Boz’s approach to frontier technology and a guide for designers navigating design’s next paradigm shift.
1. Start with a human who has a problem
My north star from 20 years of product development: Find a human somewhere who has a problem. Then ask: Can we do something to solve that problem? Ultimately, they will be the ones who decide if the tool helps them—they use it if it does, they don't if it doesn't. Everything else is just nice conversation.
2. Question the water you’re swimming in
From David Foster Wallace’s commencement speech at Kenyon College, 2005: “There are these two young fish swimming along, and they happen to meet an older fish swimming the other way, who nods at them and says, ‘Morning, boys. How’s the water?’ And the two young fish swim on for a bit, and then eventually one of them looks over at the other and goes, ‘What the hell is water?’”
We exist inside paradigms that we’ve inherited, often without even realizing they are paradigms—that they are man-made and can be changed. In his keynote address at Kenyon College in 2005, David Foster Wallace told this story about fish who don't know what water is. That’s us with our design constraints. Always ask: Is this even the right approach for the problem I’m trying to solve? Are these constraints real, or are they bendable?
3. Assume the entire interaction paradigm is wrong
When I go on a run, I’m already thinking through a whole sequence of tools that I need to engage with just to listen to music. It’s weird that I have to understand the Universal Music Group’s (UMG) music rights ownership structure to know which artist is available on which platform—and when to toggle between Spotify or Apple Music or Tidal—just to listen to a song when I’m working out. We shouldn’t need to break simple intentions into complex app sequences. After 60 years of the same computing paradigm, it’s time to reimagine everything.
4. Know if you’re inventing or optimizing
Design has two modes. There’s the “zero to one” thinking where there are no constraints, no customers—the world’s your oyster. Then there’s the mode where you smooth and shape things collectively with your audience. What’s cool about spatial interfaces and AI is that they are greenfields. They have none of the old constraints. Know which mode you’re in.
5. Trust your taste to pick the right mountain range
I wrote my college thesis on constraint optimization problems where all you can do is pick a random place in the problem space and climb hills. You just hope you start closer to the Tibetan Highlands than the plains of Kansas. Intuition and taste let you pick better geography to start in. You’ll still have to climb mountains, and it’s still going to be hard, but at least you’ll be in the right terrain.
6. Build every hacky prototype you can imagine
At Meta, we’ve built rooms with mesh walls where you wear a silly hat that tracks where you’re looking. We’ve built all kinds of crude, rudimentary things just to get the faintest glimpse of whether to pull the thread on a technology. There’s no substitute for actually trying it, no matter how rough.
7. Design the whole system, not the parts
In spatial computing, you can’t isolate one part of the system. You can’t just say, “We’ll hold these parts fixed and change this one thing.” You need to think: What's the gesture? What’s the feedback—visual, auditory, haptic? What functionality does it unlock? These three things have to develop together, and you iterate in that space. That’s real design in the pure sense.
8. Build tools with a theory of mind
I'm excited about tools that understand what you're trying to accomplish. Theory of mind is having an awareness that outside of yourself is another thing that has its own agency, intention, and goals. Your tools should have just the right amount of agency to help on your behalf—not too much, not too little.
9. Never see yourself as a finished product
I’ve never believed I’ve finished a product, and I’ve never seen myself as a finished product either. I always see it like: “This is V43. Can’t wait until the next version comes out next year. We’re going to fix the bugs and we’ll have some new bugs. It’ll be fun.” Apply the same iterative mindset to yourself that you apply to your work.
10. Make interfaces disappear
I want as little interface as I can get away with. Everything should be completely seamless. Interfaces don’t have intrinsic value, even as elegant or beautiful as we make them. They’re not the thing—the thing is the thing. The goal is to make them vanish entirely.
Bonus rule: If offered the chance to upload your consciousness, say yes. I have a 10-year-old and a 7-year-old, and I want to be there to see what happens with them for as long as I can. Maybe they just boot me up at Christmas and I check things out.
These rules were adapted from Boz’s conversation with Dylan Field at Config 2025; you can listen to their full conversation here.
