What I learned from my own years building a virtual world
“THIS IS GOING TO BE THE 3D INTERNET” a tech reporter boldly wrote. When was this claim made? Not in a 2021 tweet covering Mark Zuckerberg’s vision of the metaverse, but in an early 2000’s magazine article forecasting a rosy future for the virtual world Second Life (SL). I was working there at the time, a young team member in his first startup post-grad school. As the ‘non-engineer,’ I did a little of a lot, and a lot of a little: consistently some product management work, but also tasks as diverse as writing our first community standards, negotiating the software license for a physics engine and trying to forecast when a $400 out-of-the-box Dell tower PC would be able to meet our minimum hardware spec. You emerge from experiences like these with [Liam Neeson voice] a very particular set of skills, and I’d say my nearly three years at Linden Lab, the parent company of Second Life, certainly impacted all the work I’ve done since.
Some of you might be familiar with SL but here’s the short story: there’s a historical category of ‘metaverse’ products that tend to gravitate towards the user-designed, open-ended, physical simulation attribute set. They can be centralized (‘land’ is hosted by a single company) or decentralized (anyone can host a server so to speak). Second Life was one of the most notable early entrants in this category, with ambitions and innovations which transcended its perpetually plateaued consumer adoption. Minecraft and Roblox are infinitely more successful versions of Second Life with the ‘realism’ attribute dialed down.
All of this is to say, you can imagine that as our industry “pivots to metaverse” has led different people to ask me, what did you learn from Second Life? I’ll tell you what I learned. There are going to be toilets there.
The question about whether your virtual world should have toilets is fiercely debated. Ok, maybe not. But I’ll tell you, watching our users build bathrooms and toilets in a world where avatars don’t need to pee or poop (let alone have genitalia. Yes, it’s true that SL avatars had attach points on/near the waist where maybe it was possible users designed performative or ceremonial genitalia but the natural form did not), was surprising. But it goes towards an understanding of what artifacts we carry forward from our banal humanity into these virtual spaces and which do we seek to shed. It’s why metaverse platform builders need anthropologists not just engineers.
I’m using this short toilet anecdote mostly as a way to share the work of Nick Yee, who impressed me with his research and surveys of massively multiplayer online games (MMOGs). I hope it doesn’t get lost just because it’s ‘old’ (2002–2009). Nick’s work confirmed for me why and when people bring IRL habits and etiquette into simulated environments (for example, avatars tend to stand around in clusters that mimic the spacing and positioning that we’d take in our own conversations even if it’s unnecessary in a digital world).
Technologists have a tendency to always be looking ahead, sometimes with a bit of optimism or naïveté (or ego) that what they’re doing is uniquely new. I’d never want to remove this completely because as a whole it serves us well. But Beginner’s Eyes tempered with a bit of compounding learnings serves us even better. So I hope in the rush of new talent heading towards metaverse projects we don’t forget to distill what we can from the past.
Update 12/2
The best thing about writing posts like this is they invite other memories from my Second Life colleagues. Frank Filipanits, who was there even before me, shared this:
