Talking about the metaverse is probably the least sexy thing a person can do. I had a conversation about what the metaverse is with a coworker this week, and I’ve never felt like more of a dork in my life (I work at a data science consulting firm). Unless you tell them we’re all gonna live with VR goggles glued to our skulls playing SIMS or Roblox for all of eternity, no one has a clue what the metaverse is, or why it matters. It’s even harder to have the conversation when you start trying to breaking down important topics like interoperability.
Anyone that’s connected to this space has heard of interoperability, and they’ll know that it’s one of the key foundations for what makes the metaverse so special. It’s not a piece of tech that a company like Apple or Facebook can gatekeep and hold on to. It’s kind of the opposite. Interoperability in the metaverse represents the successful connectivity of all ecosystems on the future web (or at least few). It’s one of the most important concepts surrounding the metaverse, and to me, it’ll be the barometer for testing whether or not whatever we end up creating is actually a metaverse.
Earlier in the week when I tried to explain this topic, I dove straight into the facts. It didn’t go very well. So this time, I’m gonna start off by laying out a scenario. Something simple. Something light. Something everyone will understand.
How Quantum Physics and String Theory Solve Interoperability in the Metaverse
Online avatars are a great example for explaining to someone how interoperability could work in the metaverse. Creating a digital representation of yourself is something that everyone loves to do (even if they don’t know it). That’s why RPG’s are one the best genres of gaming. They allow you to create your own character and throw it into a world that’s larger than anything you’ve ever experienced. Even if you spend 3 hours adjusting the width of your character’s left eyelid, it’s always worth it.
Scenario
Just imagine, that in some point in the future, Zuckerberg does it. He releases Horizon and the metaverse is real. When you go to sign up for Horizon, you no longer use the archaic method of uploading a profile picture, but instead, you spend your time creating a Sims like character to represent your identity. Oh, and it’s not just the physical features that it’ll represent, but the aesthetics too. Hopefully Meta’s attracted some top fashion brands like Nike or Gucci so that you can get your avatar fitted out. I can imagine a whole digital economy revolving around fashion brands selling outfits and pieces for your avatar to wear (don’t worry, for all my broke boys, I’m sure TJ Maxx will be in the metaverse too).
Now let’s say in this same future, Twitter also releases a metaverse platform — we can call it, BirdCage. All your friends ended up signing up for BirdCage when it released and now it look like they’re having way more fun over there. You feel left out. You want to join them, but you really don’t feel like spending another 3 hours creating an online avatar again. Thankfully you don’t have to. This is the metaverse, everything’s connected. So you open up BirdCage’s sign-up page and see the “import avatar” button there. You click it, and 5 minutes later, your avatar from Horizon is fully ported over. You get a big smile on your face :), you look to sky and scream “Interoperability is the best!”, and sail off into the metaverse living happily ever after.
Well that is until you actually SEE what your Horizon avatar looks like in BirdCage. You quickly realize that the clean and crisp avatar you spent hours (and 10’s of $’s) creating on Horizon, now looks like a Minecraft character that shops at goodwill.
What the heck happened? Well, interoperability is what happened.
No seriously, what happened to my character???
Sending an avatar from Meta to Twitter’s metaverse platform seems like it’d be a pretty simple task, but it takes a lot of work, and I think it encapsulates all the basic principles of interoperability.
Meta has to be able to take your avatar, convert it into some readable data format, and then send that file over to Twitter. Twitter then has to showcase the capability of not just reading the data file, but also interpreting it, relative to its native environment, and recreating an object out of it. In the scenario I gave, the end result wasn’t perfect, but the mechanisms at play still captured the essence of interoperability (at a basic level).
So if the essence of interoperability was maintained, why did BirdCage turn our dripped out Horizon avatar in a 420p meme character? To put it simply, BirdCage didn’t now how to interpret the data correctly. It may have been able to read all the 0’s and 1’s, but executing the correct procedures from them, as well as rendering an object that is subjective (what it looks like depends on the environment), are really complex tasks that no one has built a great solution for yet.
The concept of automating the transfer of assets from one environment to another is the problem that interoperability hopes to solve. From the example I just shared, it seems like the solution is just to build out applications and tools that are better able to interpret data sent from foreign environments. This is definitely a great solution, but it’s not the only one.
To me, there are two different ways we can see interoperability come to life in the metaverse space: tech and strategy. In the future, I think both will coexist and power and interoperable metaverse that allows ecosystems to grow and be connected in a way that’s never been seen before. But for today… strategy wins.
Interoperability Implemented
I think the diagram I spent 17 minutes creating gives a pretty good representation of how interoperability can be realized in the metaverse through time. In the beginning phase, I just don’t believe there’s enough infrastructure built around converting assets across ecosystems for high quality technical solutions to be present. As time moves forward though, more and more frameworks for development will be established, tools will get built out, and we’ll start to see the emergence of interoperability from the technical end of the spectrum.
Right now I think strategy is where companies need to focus their resources for creating interoperability.
Partnerships
Utilizing partnerships is the simplest solution. It’s also the easiest (at least for your engineering teams). High value brands today are just as fired up about the metaverse as the rest of you dorks reading this article right now (don’t worry I’m king dork for writing it), so it’d really be a waste to not take advantage of the potential opportunity.
A great example of a brand-platform partnership happened at the end of last year when history was made by Steph Curry. Back in December 2021, Under Armour partnered with Decentraland (a crypto metaverse platform) to release NFT wearable Steph Curry sneakers in honour of him breaking the all time NBA 3-point record. Now if I’m being honest, this alone isn’t that COOL. We’ve seen TONS of celebrities do collaborations with different brands to release some kind NFT collection that they can cash out on. That by itself isn’t groundbreaking anymore. However, what IS COOL is that these wearable NFT sneakers weren’t going to be released exclusively on Decentraland. They also were gonna be dropped on 2 other crypto metaverse platforms: The Sandbox and Gala Games. Now the REALLY COOL part of ALL of this is THAT (alright I’ll stop) the NFT (doesn’t count) sneakers can be worn across all 3 platforms.
This was accomplished through coordination. Not advanced tech, but just some men and women on the phone chatting it up. Gala Games and The Sandbox don’t have some advanced technology that allows them to automatically recreate an object that isn’t native to their environment. However, what they do have is business development team ready to reach out to brands and have them build cool collections on their platform. I’m positive all 3 of these platforms worked together so that their individual game engines would recognize the asset from the other two’s platform and create their local version of the sneakers in its place.
It’s these kinds of relationships that will give rise to interoperable environments in the near future. Even when the tech to fully recreate foreign objects becomes available, I’m sure partnerships will still have a huge part to play.
Open Source Development
In this day and age, open source libraries and tools are the backbone of the software development community. Seeing how the metaverse is being built by software developers… I declare that by the transitive property, open source tools will be the backbone of the metaverse.
Now the math there doesn’t really add up, but, it’s not a crazy claim to believe open source development will propel the metaverse towards interoperability. I mean, one of the largest drivers for the metaverse is blockchain, and the blockchain developer world is massively open source (don’t make me whip out the transitive property again).
Now I keep on saying open source, but I know that it doesn’t really mean much if I don’t give any context. What I’m not talking about is a company pushing their massive codebase onto GitHub and hoping some nerds contribute to their project. I’m making the bet that metaverse technology is so new, and that open source development is so established, that from the beginning phases of development, the larger community will dictate how the metaverse is created.
This would lead to more projects drawing from the same “source of truth” when building their foundations and could even lead to similar frameworks of development forming. Skip ahead a few years and I’m sure massive libraries/packages will be built that the majority of developers in the community rely on. When this happens, you’ll eventually end up with a landscape where a ton of projects have a lot of cross compatibility simply because they were built using the same tools.
Now what I just described sounds good on a high level, but the concept definitely breaks down when there are enough low level operations happening that it abstracts from the procedure to create the “correct” end-result. Basically, just because two things are built on the same “platform” or with similar tools, doesn’t mean they know how to communicate with each other. I think this solution makes the process of enabling interoperability easier, but it won’t solve the problem explicitly.
Standardized Intermediates
So I've been pondering over the solution for this section for days now. I think I may have been a bit naïve about how to build this solution. The difficulty of interoperability isn’t with figuring out the conversion of assets (it can be done manually), what’s difficult is trying to do this efficiently (with automation).
My initial idea was to create a kind of standardized index for aiding in the process of interpreting foreign data files. I wanted to create an index that would act like a central library for understanding the architecture of specific objects from specific entities (Meta, Twitter, Decentraland, etc.), but this wouldn’t actually solve the issue of automated interpretation, it would just make the process a little more organized.
Differing levels of complexity, in terms of how objects are constructed, are the real problem here. Automated interpreters won’t know how to convert a more complex object into a simpler one (and vice versa) without having hard coded conditions in place. The only way to make this work is to create some kind standardized file format for writing the data, while also utilizing a standardized object architecture for creating the asset. This would force the data to be in a predictable enough format that interpreters would be able parse and interpret the data effectively. This solution would force a ton of constraints on the architecture of creating assets. I’m extremely hesitant to do this because its telling creators what their limitations are before allowing them to ever create anything.
So for now, I have to hold this L and come up with a better solution.
On the bright side, at least I had a third technical solution for bringing interoperability to the metaverse.
Artificial Intelligence
Don’t do it.
Thanks for Reading
