Interlude: Let's Go to the Metaverse
Or when all your open tabs keep SCREAMING at you to process them.
I already started typing the next post for my AI learning journey — but I simply have to take care of all those open tabs. Especially since the whole metaverse narrative may turn into a special project for me in the near-ish future.
So what do I have for you — and can I somehow circle back to AI while I am trying to make sense of it all?
Metaverse And Big Tech
When I was googling “metaverse business network”, I found this article on WIRED.com: The Metaverse Is Simply Big Tech, but Bigger.What I find interesting about this article is the attempt to construct a linear evolution from “isolated cyberspaces”, like websites, made by many small user entities and scattered arount the Internet, to a few Big Tech companies that consolidated many of those entities (like Facebook aka Meta acquiring Instagram and WhatsApp) — or connected them, e.g., through Google search: “For nearly 30 years, the gravity of consolidation has pulled cyberspace together under the auspices of fewer and fewer corporate titans,”
Imagine an hourglass:
At the beginning, you had all those cyberspace enthusiasts with their own digital projects. And now we are at the bottleneck of the hourglass, with Big Tech controlling many or most of the services and tools we use online. And Big Tech uses proprietary solutions — which is a hindrance if you want to develop a true metaverse:
A metaverse must be interoperable; digital services associated with it must piece together, quilt-like, to form its fabric. Matthew Ball, a venture capitalist who has written frequently on the metaverse, says, “Interoperation effectively requires companies to release their control over proprietary formats, or otherwise adopt wholly open source ones.
In the early 2000s, a bloom of open source metaverse projects emerged to solve the problem of stitching together existing virtual worlds. (…) And had the internet stayed frozen in its early form, one could easily imagine the porous, egalitarian metaverse it would have spawned: A 50-year-old in a Barbie avatar walks straight from her Second Life Dream House to Sephora.com’s VR boutique, where she purchases digital mascara with gold earned in World of Warcraft.
But those open metaverse projects never got off the ground. “There wasn’t a lot of enthusiasm around interconnection, partly because there really wasn’t a motive for it,” says Philip Rosedale, founder of Second Life publisher Linden Lab. “We were, as a company, trying to make some money.”
It seems that as of today, we are stuck with the idea of a metaverse that will be created by a few Big Tech firms:“This is the internet the metaverse inherited. Or more precisely, this is the internet Big Tech’s metaverse describes.” And those ideas are built on proprietary systems which we, the users, are only supposed to use within the boundaries of their frameworks (which I learned about in la sociologie des usages technologiques back then in France).
Note: I am not judging all of this, I am just making observations.
And if we look at the ideas Big Tech has for that metaverse, they are all based on corporate interest and also… well, incestuous. There is nothing truly new to anything I have heard about so far (no, I haven’t read everything about every Big Tech company’s vision for the metaverse). I think — and I hope I am not alone with that thought — that a true metaverse would need a larger group of innovators and inventors from all kinds of different backgrounds and with all kinds of different objectives. Where does true art have a place in Meta’s metaverse? Imagine what the initiators of Burning Man could do if they had access to some “open source” metaverse SDK — imagine going to a metaverse Burning Man.
And yes, giving everyone access to the play field will mean that there will be niche stuff like the early HTML websites with marquee text and blinking headlines in Comic Sans. But would something like WordPress have been possible without those predecessors? Would something like WordPress have emerged if everything had been Google and Amazon and Meta from the beginning? I think not.
Another thing I am concerned with is the fact that the Internet as we know it is so convenient. As a user, you are not per se challenged to dig deeper, to find out how the technology behind the GUI works, or to contribute something new to the cyberspaces we use today (on the surface, that is, I am not talking about the Deep Web or anything the like — but imagine a Deep Metaverse, how thrilling would that be?!). Yes, if you are a tactitian (as defined by Michel de Certeau), or a tinkerer (as defined by John Dewey), or a creative person (as defined by Mihaly Csikszentmihalyi), you will have that drive to look behind the curtains. But it will be very hard for you to be visible with your ideas if you don’t instrumentalize Big Tech networks (e.g., become a TikTok legend).
I observe something similar when I look at indie game developers. Once they start reaching out to build their communities (which they have to do during the game development already — own Discord server, devlog blogs, streaming, social media and all), something changes in their development process; and maybe something about the games they produce changes as well. They optimize their products for display on Steam or in the Epic Games store, they make an effort to port their games to all the devices you can imagine (because all devices use proprietary frameworks instead of fostering compatibility so devs don’t have such a hard time to come out on all platforms!) — and many of them realize that you can’t make any real money with a small indie game, even if it is a gem because of its art style, story, game mechanics or other features.
There is no place for tinkering in your garage and stunning everyone with the finished product earning a ton of money because of the excess supply of games, each of them being its very own piece of isolated cyberspace — and because of the dominance of Big Game firms with enough leverage (aka marketing budget) to just occupy the time and attention of huge and global audiences.
Now, imagine the games industry without Big Game firms. You’d have a billion small game developers, and a certain number of them would recognize the benefits or working together on open platforms in order to create comprehensive cross-games experiences, walking from one game to the next and connecting them to a big meta game.
This vision may seem a tad utopian but if we look at open source projects, I think we see the potential lying therein.
So if we are at the bottleneck of the hourglass right now with Big Tech (or Big Game, if you prefer that) that consolidated everything to the maximum while depriving us of the agency to co-create the future metaverse, we can only hope that a counter-movement will be initiated by those advocating an open metaverse — one that does not stem from a few large corporations but from the creative energy of the masses playing around with new-ish technologies. It doesn’t all have to be perfect at the beginning. Just give us the open space to experiment ourselves.
Metaverse And Big Data
Another concern that needs to be raised — or a question that needs to be asked — is this: What about all the data? And that implies multiple paths we can follow, such as these two:
What about the data Big Tech will gather about me across even more platforms (like cookies but a bit more sophisticated)?
What about processing or managing — let alone making sense of all the data that will be produced?
There are definitely more questions and concerns that could be added to this list but I do want to finish that post at some point today, so I have to stop somewhere.
Let’s start with 1. What about the data Big Tech will gather about me across even more platforms (like cookies but a bit more sophisticated)?
Well, yeah. Part of the idea to enable a myriad of smaller actors for creating a metaverse is that collected data with be a lot more scattered instead of being consolidated. I am not the kind of person who freaks out when she sees ads for REBELLE pre-loved designer bags on a website after she has browsed the website for far longer than she wants to admit (ah, the cliché!); and I am not concerned when Facebook and Instagram exchange information about my browsing habits. More often than not, I find the recommendation function on Amazon useful (but let’s not talk about the behemoth of Amazon product listings, seriously, why can’t someone just clean that mess up?!) — and I love my Discover Weekly playlist on Spotify. I do however apply selective usage patterns. I’ve never used any Facebook reaction other than the standard “Like”, I don’t heart songs on Spotify, I just add them to a super random playlist (this one) and I deleted all my Amazon wishlists.
(2 minutes break to verify I really did delete them all… except for the one with paperback books I gave away where I am hoping there will be Kindle versions at some point in the future)
I am not following these self-made rules to “trick” the algorithms. I am not even sure if that actually impacts the way how data about me is gathered — or if data regarding which reaction I show on Facebook is only accumulated without being put to use (see 2.). And the reason why I am buying from 3947537459 different online shopsis not to keep Amazon or eBay from knowing everything about my purchases. It’s simply because I want to get the best bang for my buck — and Amazon and eBay do NOT have everything.
So really, there is no strategy behind my behavior except I am indeed hoping I won’t mess up the delightfully chaotic and oftentimes shitty “Discovery Weekly” playlist Spotify throws at me if I don’t tell the system too much about what I like, so that maybe my bubble (be it on Spotify, Facebook, Twitter, or Instagram) is maybe not as restrictive as someone else’s who discloses more about themselves by using all the functions as intended by the developers.
And I am thinking about how I would want my data to be collected when I use something like the metaverse. I think any universe would be a lot less messy if every product had just one instance of itself instead of thousands of duplicates. If you look for the standard Continental 5000 road bike tyre, you can find it on thousands of websites with dedicated product IDs and descriptions (mostly copy-and-paste from the original Conti description). You can find 209 instances of exactly this item on eBay Germany alone — instead of just one with the different options (23-622, 25-622…) and the different pricings for all the different web shops in the wild (like Google Shopping already attempts to do but of course, Google Shopping only features instances of the item that Google can or wants to find). If I look at the beginning of the paragraph, I can’t quite say how unique product IDs are related to my data being gathered in the metaverse… but I do think that a metaverse environment could help declutter the Internet (and save online shop operators so much time) by interconnecting all instances of one product behind one product ID (STORED ON THE BLOCKCHAIN! Ultimate win!) — with a global login.
Ah, now I remember what I wanted to say!
As much as I like shopping in different online shops, it is different from going shopping in different stores in real life. I leave my data behind in every online shop I visit — even if many allow you to use Amazon or PayPal directly “without signing up” (because your payment provider will do that for you in the background…) and even if you can order as a guest at many of them (you still get an ID associated with your purchase, you still leave your email and real-life address for the delivery). If I buy a piece of cake in a bakery with cash, the bakery knows nothing about me (except maybe that I tend to show up there not just every once in a while because I love cake).
That’s where a metaverse would come in handy as well. I would love to be able to ditch all the sign-ups and logins (after all, those are all avatars of yourself that you leave behind) if purchases could be ascribed to, well, that one instance of me that really is me, not an avatar. My unique user ID as a human.
But as with all technologies, it is only as good as its acceptance rate in the public. And especially as a German person (many Germans are very concerned with privacy), I can already anticipate the concerns people will raise — which could be curbed by scattering one’s own data traces, i.e., by using many different small services instead of Big Tech. Of course, this mostly makes sense for online shopping and other non-social activities. Social networks are a completely different thing to look at.
One last thing about data being gathered: I would argue that data collected in a metaverse is both more and more precise than data gathered from regular Internet usage — essentially because everything wants to be connected. But this doesn’t have to be a bad thing if anonymization (YES, I AM LOOKING AT YOU AGAIN, BLOCKCHAIN) and choice can be ensured. But choice means education — so everyone would have to be aware of what their actions mean in the metaverse, and where their data is gathered to what extent. And this is a highly utopianist stance. As easy as using Internet applications may be (and it really isn’t that easy or obvious for every netizen either), it is really hard to grasp the implications of every single action, or to consciously make reflected choices about one’s own usage. How many people at Meta, Alphabet, or Amazon (I can’t wait for the moment when Amazon decides to rename itself as well) do actually understand the algorithms at play? I am guessing there is only one viable solution to all of this, and it would be — yeah, guess what, I made the connection (!) — AI.
Which leads me to 2. What about processing or managing — let alone making sense of all the data that will be produced?
I work at a rather small initiative at the intersection between the games industry, politics, media & arts, and the public. But even we have more data than we can process. Some metrics would be: analytics for our website (and in case you haven’t heard it yet: since the new GDPR and cookie regulations, website visitors are nothing but wild guesses by any analytics tool — you just can’t track your cohorts or user groups any more like it was possible in the early 2010s), following and engagements on our social media platforms, leads from all kinds of industries with their specific needs, and some more. I can’t keep track of all of that. And if you are not a dedicated professional whose job it is to process, manage, keep track or make sense of data, there is no way you can catch up with the data as it is being produced. There may even be noeuds in your system where you could collect even more data which you are not aware of because you don’t have the time to start looking for them.
So — with the limited knowledge I have about AI so far — I suppose AI would be the only solution to the problem of keeping up with data. Of course, such an AI (such AIs?) have to be trained for this purpose but I don’t see that as the problem here. And an AI could come up with new readings for data as well that could make gathering ALL THE DATA obsolete at some point — maybe because it would develop a sophisticated set of X personas the behaviors of which it could predict based on historical data and based on a limited amount of new data coming in (e.g., in times of a significant crisis like COVID-19 where people may behave differently since everyone is going nuts after two years of pandemic mismanagement).
This would further contribute to alleviating privacy concerns. Imagine your business model worked and thrived without having to gather all the data from everyone at all times?! Less data garbage, less energy consumption, less storage space (and devices) — and at the same time, more intelligent decisions. I even think that not knowing everything (which is also just the illusion of thinking you could know everything about your customers) would be beneficial for businesses. If I look at advertisement today, I cannot but SMH in disbelief since none of the modern ads in general mass media appeal to me. Internet advertisement has never not sucked in the first place, but I remember some good ads from an era where creative decisions where not predominantly based on what data tells you your customers might like. I am not saying businesses should not know their customers — and I am not saying extensive data analysis is not important in many B2B scenarios — but too much data can kill innovative and creative processes, so a little less of the compulsive analyze-everything-before-doing-anything might be a good thing.
At the same time, I cannot but SMH (yeah, again) at how dumb so many tools are. I love Google Workspace but why can’t I move items from Drive in one browser tab to Drive in another browser tab? Why is it so cumbersome to move items or to navigate the folder tree? Aren’t you all using Windows or MacOS and don’t you all know what competitors like Dropbox are capable of? Why can’t I copy something from one calendar to another calendar without switching to the Google account hosting that calendar? This is not the fluid workflow I expect from a dedicated workspace environment that should assist in transitioning to a metaverse. Yes, I know that Google hasn’t jumped on the metaverse train yet but if the interconnectedness of everything is the next step in our technological evolution, Google can’t stay out of it. And we know there are things going on at Google towards AR — which I would say depends even more on flawless and soft transitions between tools if you don’t operate your AR activities with a mouse and a keyboard.
Or maybe that’s just me being sad about half-finished products that are abandoned because there’s another hype train passing by you have to jump on in order to stay relevant.
Metaverses and Where to Find Them
As for every topic, there is a shootload of resources out there in the wild and it is easy to get lost while trying to find something that could potentially contribute to a metaverse. One could argue that a movie franchise (like Marvel or DC) with several iterations, websites and fan wikis, merchandise, websites, and even games featuring the heroes might form some sort of metaverse — but those different cyberspaces are not connected, so that doesn’t quite count.
If you ask me (I know you aren’t, but you are still reading my blog, right?), the most metaverse-ish applications we have today are:
Second Life (2003), a connected and user-defined world — and yes, it is still up and running
Roblox (2006), a platform of user-created games — but not interconnected
Minecraft (2011), basically similar to Roblox — but with less (no?) opportunities to make money from it within its own ecosystem
Fortnite (2017/2018), with the Creative mode reminding you of Roblox — but also huge in-game events like concerts on the dedicated Battle Royale map(s)
Can you see a pattern emerging here? Yes, those are all platforms with a heavy focus on UGC and creative freedom. And yes, those are all games. I am not saying gaming companies should build the metaverse, but maybe they should. If you bring together all the creative potential in all the game designers and developers from the entire world (minus the ones copy-and-pasting established addictive game mechanics for the sole purpose of milking their clients, sorry if I sound a little bitter here), I think people would come up with something better than the trailer Meta gave us. Furthermore, I am convinced that a metaverse needs to be fun to succeed. It needs to be rewarding and gamified in all the ways you can and can’t (yet) imagine.
Which brings me back to the idea that AI (an open AI framework!) would be necessary to coordinate and co-create such a gigantic thing.
But apart from the games I listed, I found a project on GitHub named Awesome Metaverse that lists a few technical entry points (e.g., WebXR and protocols/standards) but also popcultural references and white papers. Changes can be applied via pull request. I am very tempted to copy-and-paste the entire list into my post but I am sure it won’t be deleted anytime soon, so just head over and see it for yourself:
High Fidelity has a Parts List for the Metaverse from 2019, posted by Philip Rosedale, the “maker” of Second Life:
While we may not yet know exactly what the Metaverse will look like, if we assume it to be a very big digital world with lots of people running around doing interesting things together, you can build a rough list of the key parts required beyond VR hardware and a fast internet connection:
I especially like what he writes about “Programmable Atoms”, since that refers to our currently limited computing power — so items cannot be broken down into units as small as physical atoms due to the restrictions, and digital atoms must be bigger than their physical counterparts.
A much older but still relevant article is A Maker’s Guide to the Metaverse from 2015 by AI researcher Rod Furlan on SingularityHub. He formulates a few tenets he considers important for a real-world metaverse:
Tenet #1 – Creative freedom is not negotiable
Tenet #2 – Technological freedom is not negotiable either
Tenet #3 – Dismantle the wall between creators and users
Tenet #4 – Support for worlds of unprecedented scale
Tenet #5 – Support for nomadic computation
He provides explanations for all these tenets and the fifth is the most thought-provoking one: “In the ideal metaverse, a nomadic program is a fully autonomous participant with the similar ‘rights’ of a human user. Like any ordinary user, such programs can move from one server to the next on the network. To the underlying computational fabric, there is no meaningful distinction between human operators and nomadic programs other than the fact that programs carry along their source code and internal state as they migrate to a new server.”
I find this thought thrilling from a personal and from a sociological perspective. We have been trying to tell machines from people apart in real life and in fiction (e.g., androids, Turing test, chat bots, Westworld…) — but what if that distinction simply doesn’t matter for the metaverse? It does of course matter for human-human and human-machine interaction, but AI could be a valuable companion in operating the metaverse if it is not bound to servers or networks. Of course, I have no idea how much of all that is already possible and of course, I would instantly devour any dystopian cyberpunk novel turning this idea into a destructive scenario without a happy ending for makind. But it still seems like a thought worth thinking.
Turns out this whole post wasn’t so much an interlude but another step on my AI learning journey. I am delighted. Thank you for sticking with me!
Thanks for reading Thys.ai! Subscribe for free to receive new posts and support my work 👇
Just a short side note on WIRED: If you know a thing or two about the history of WIRED, you should know that it seems to have been the commercially more successful alternative to MONDO 2000, which is sometimes labelled the “Whole Earth Catalogue of the 1990s”. The reason why I keep talking about those pre-cyberculture and cyberpunk artifacts is simply because some kind of “metaverse” was already envisioned back then. It just sometimes included nature as well, a more hippie-esque version of a global interconnectedness. We don’t see that in the dystopian visions of cyberpunk, but I am getting some similar vibes from the increased efforts to do something “against” climate change, so everything nature is claiming back some of the relevance it lost due to hyper-capitalism.
Maybe you remember the link to that super wild Quora question I posted last time to make a point about why AI perception in the public can tilt towards negative. I still had that tab open (“Is AI learning a scam”) and the respondent had posted a chart of the tech companies involved in the AI “conspiracy” (it reminded me of that one but without the fancy interconnectedness). Anyways — when I looked at the companies that chart lists, I thought, wait a second, you can take all those companies and put them on a chart about the metaverse as well.
This article seeks to explain the genesis of the sociology of using ICT in France of the 1980s: https://hal.archives-ouvertes.fr/hal-01679444/document — and this one talks about the usager (user): https://hal.archives-ouvertes.fr/hal-01611076/document — but this one is the most pertinent one, I’d say, since it is a critical inspection of this “sociology” (and was written by one of the professors at my university in Paris, I am so biased): https://www.persee.fr/doc/reso_0751-7971_2000_num_18_100_2235
I actually never thought about that until a friend, B., mentioned it in November 2020: What, you bought stuff from YET ANOTHER online shop?! How do you even keep track of all those shops? — Well, if you buy specific things (e.g., low-carb cheesecake), you will only get them in specific shops. Just like in real life. And if shop A has a better offer than shop B although both are cycling shops, you buy from shop A. And next time, you’ll maybe choose shop B.
And Epic Games CEO Tim Sweeney is fighting for an open metaverse (?): Be patient. The Metaverse will come. And it will be open.