From my experience playing social VR, I think Facebook Meta has it all wrong. If you go around VRChat communities you'll find an overwhelming amount of anime girls and furries make up the user base and creator base. In VR, your avatar isn't just a little picture next to your name, it's your body and your identity. Some people will pay hundreds or thousands for an artist to make a custom and unique avatar for them, or they'll spend hundreds of hours learning and making stuff themselves. Some people like to be a robot, a hologram, a pooltoy, some people like to have four arms or be a centaur, there's different styles and body proportions, practically anything is possible and creativity is the only limit. There is a whole economy of avatar asset creators on sites like booth and gumroad, and there are artists who make a livelihood off of doing 3D avatar commissions.
Facebook Meta is basically trying to throw all of this away and pretend none of it exists, and force people into one-size-fits-all humanoid avatars that people can customize slightly with some sliders and options. In fact they weren't even sure they could trust people to have legs. It effectively simulates the real-life feeling of having body dysphoria and not being able to change your body. It's virtual reality for christ's sake! There's no physical reason why somebody couldn't be a Blender default cube if they wished...
Facebook Meta is trying to make a sterile platform safe for business meetings and advertisers, while simultaneously trying to attract users and creators so that it's not a ghost town and has actual things to do, while still trying to retain a monopoly on avatar customization so that they can eventually earn massive wealth by selling virtual clothing for $10. The result seems to be a massive waste of money that ends up catering to nobody (except to Zuckerberg?). It feels a bit like if they made a computer in the early days of computing that didn't support programming, just because they hope if this nifty "computing" idea ever takes off they'll have a monopoly on programming
I was at an event where people were making these avatars, many first time users. One person who gave feedback at the end said he was frustrated he could not get the avatar to look like him.
I think whatever the hell Meta is doing with their weird alien humanoids is far from "normie" appeal as well. The furries and anime girls seem more normie in comparison.
I don't know if there is some middle ground that would actually be appealing to some definition of a "billion normie"s. Maybe actually photorealistic looking humans? Making the graphics not look like it's from 2003? Or going the other way: make them look like the Mii characters with Nintendo? Something totally different? Maybe appealing to the furries and anime girls would be actually a good idea at first, to build up some "power users" or whatever, and then attract more casual users.
I share the sentiment of the Instagram users in the article and the grandparent; it is baffling to me why the product looks so terrible, with so many resources poured into the Metaverse.
> Almost as astonishingly, revenues at the division have actually fallen from the 2021 high of $2.3bn, to $2.1bn in 2024...
Not sure why this is astonishing. It was new back in 2021. People tried, mostly hated it, and moved on. Honestly, I'd consider the $200mn drop a victory.
It depends how good the AI is and how it's used etc. It would for example be neat to walk around in a massive AI generated multiplayer city that has a lot of intricate details.
Humans are really good at identifying the unreality of procgen because it loses the intentionality we expect from our built environments. Here's the author of that game talking about their solutions to the problem:
Procedural generation isn't what we call "AI" now, but AI is absolutely procedural generation.
If your inputs to an LLM are prompt + seed, isn't that analogous to world config + seed in Minecraft? In fact, I would argue they both have the same problem: In an infinite world, geography has no meaning.
Why do I get bored of exploring new minecraft worlds? Because ultimately, I know the limits. I've seen cool minecraft mountains, what do I care if this time they're next to a desert but last time they were next to a lake? None of it has any meaning, there's no context. That mountain isn't a holy site for some religion, or a reminder of an ancient battle. It isn't a clever ploy by the writers of a game to set the stage for a final fight. No, it's just mountain #9057382 in minecraft world #74893124.
Midjourney's "wizard tower" with seed 1 vs seed 2? There are differences but none of them mean anything.
I played with AI image generators for a whole month and then I got bored. Not because I can't think of other prompts, but because "witch cat riding a broomstick" is something I can easily imagine in my head and that takes about the same level of effort as typing it into midjourney. And the result has the same meaning. None of my friends want to see "witch cat riding a broomstick" #A213DEF675, the novelty is gone. If they wanted to see a cat on a broomstick they could type the same prompt in.
Now, I only use image generators for D&D campaign art, and I write the campaign myself. Because I know my friends, I know what they like. I know what references they will understand, and every detail I hide in the world is something special for them.
I don't see the "metaverse" being any different. It's cool that we figured out a way to make the monkeys on the typewriters turn out stuff that is readable language instead of gibberish, but how can any work ever stand out in an infinite library?
Dead Cells uses procedural generation but does it in bespoke chunks, so while a level is different it’ll have chunks that are determined fun and then glued together by the procedural algorithm.
Yes, all of the games with the best procedural generation achieve it by heavily leveraging handcrafted pieces, but that's an indictment of procedural generation, not an endorsement.
Sorta? Procedurally generated levels tend to not invite repeated playthroughs. Maybe this is a leap above previous efforts, but also maybe that isn't a problem.
I'm all for folks trying. I'm less enthusiastic it will pay off.
> Procedurally generated levels tend to not invite repeated playthroughs.
I agree with your broader point about low confidence in an AI-powered quantum leap, but in my experience this point about repeated playthroughs is actually backwards. Lots of the most popular and/or heralded games in recent years use procedural generation to enable repeated playthroughs as a core mechanic. Balatro, Hades, Slay the Spire, Diablo, Helldivers, Civilization, etc etc. Even things like drafting in MtG or board layouts in Settlers of Catan use procedural generation to increase replayability.
When the breadth of the content is just about infinite, will they care about repeated playthroughs? I think they could capture quite a few people just keen to explore and see what's around each successive corner.
I think they could do things like check the response to some areas and favour those to weed others out.
I remember spending time in RDR2 exploring the landscape, and finding the variety of moments that spawned (hunters, animal attacks, peril and whatever else). I think AI can get close enough to that quality of environment to satisfy people. As it was, the level designers were clearly inspired by the natural world which was not generated by human game designers.
“Metaverse” is a moving target and has one goal for Meta: get everyone off the platforms that can eat their lunch (Chrome, iOS, Android) and onto a platform where Meta can start collecting 30% from everyone who wants to play.
People don't want to interact with the human-generated content, why would AI-generated content change their minds? I have yet to meet anyone who uses the metaverse, and it's not for lack of funding on Meta's part.
That’s because you aren’t 12 years old. Ryan George (a YouTuber) made a video where he checked out the metaverse and it was basically exclusively filled with children. He played metaverse worlds for around 5 hours straight and didn’t meet a single adult the whole time. It’s children using their parent’s old Meta Quest 2 headsets they bought during the pandemic because they were bored, then forgot about.
Until I locked my headset down my kid got onto the metaverse a few times without me realizing, and playing a bunch of random “experiences”.
The few times I (50s male) tried horizons universe I felt a visceral creepiness - that any adult I “met” was likely there because it wasn’t technically within 100meters of a school.
VR Chat was less locked down and felt more self patrolled and honest. (On the other hand I wandered a common “watch a movie” vrchat video room instance where Nazi propaganda played and the people in the room were bro-asting about being racist anti-semites.)
Anything that can generate metaverse activities can do the same for flat-screens, which meta has been failing to convince people away from for many years despite billions of dollars of investment.
You can't blame them for trying though. Their cash cows aren't going to last forever. Actually, I don't understand why this company has such a high valuation. From all the FAANG, it seems to be the less diversified.
Metas ad platform is unfathomably large. Their tracking capabilities on mobile make them the ad market most apps use to serve ads. Apple tried to kill that but turns out it wasnt so easy and meta came back as a better tracker than ever.
https://archive.ph/DQWhw
From my experience playing social VR, I think Facebook Meta has it all wrong. If you go around VRChat communities you'll find an overwhelming amount of anime girls and furries make up the user base and creator base. In VR, your avatar isn't just a little picture next to your name, it's your body and your identity. Some people will pay hundreds or thousands for an artist to make a custom and unique avatar for them, or they'll spend hundreds of hours learning and making stuff themselves. Some people like to be a robot, a hologram, a pooltoy, some people like to have four arms or be a centaur, there's different styles and body proportions, practically anything is possible and creativity is the only limit. There is a whole economy of avatar asset creators on sites like booth and gumroad, and there are artists who make a livelihood off of doing 3D avatar commissions.
Facebook Meta is basically trying to throw all of this away and pretend none of it exists, and force people into one-size-fits-all humanoid avatars that people can customize slightly with some sliders and options. In fact they weren't even sure they could trust people to have legs. It effectively simulates the real-life feeling of having body dysphoria and not being able to change your body. It's virtual reality for christ's sake! There's no physical reason why somebody couldn't be a Blender default cube if they wished...
Facebook Meta is trying to make a sterile platform safe for business meetings and advertisers, while simultaneously trying to attract users and creators so that it's not a ghost town and has actual things to do, while still trying to retain a monopoly on avatar customization so that they can eventually earn massive wealth by selling virtual clothing for $10. The result seems to be a massive waste of money that ends up catering to nobody (except to Zuckerberg?). It feels a bit like if they made a computer in the early days of computing that didn't support programming, just because they hope if this nifty "computing" idea ever takes off they'll have a monopoly on programming
Meta needs VR to hit big with normies and what you describe will never hit with a billion normies. So it's not surprising they aren't focusing on it.
I was at an event where people were making these avatars, many first time users. One person who gave feedback at the end said he was frustrated he could not get the avatar to look like him.
I think whatever the hell Meta is doing with their weird alien humanoids is far from "normie" appeal as well. The furries and anime girls seem more normie in comparison.
I don't know if there is some middle ground that would actually be appealing to some definition of a "billion normie"s. Maybe actually photorealistic looking humans? Making the graphics not look like it's from 2003? Or going the other way: make them look like the Mii characters with Nintendo? Something totally different? Maybe appealing to the furries and anime girls would be actually a good idea at first, to build up some "power users" or whatever, and then attract more casual users.
I share the sentiment of the Instagram users in the article and the grandparent; it is baffling to me why the product looks so terrible, with so many resources poured into the Metaverse.
> Almost as astonishingly, revenues at the division have actually fallen from the 2021 high of $2.3bn, to $2.1bn in 2024...
Not sure why this is astonishing. It was new back in 2021. People tried, mostly hated it, and moved on. Honestly, I'd consider the $200mn drop a victory.
Did they set their expectation much higher than that? Because it sounds like too exaggerating for 1% drop
It's astonishing how much money they are pouring into a product that is just seems like a worse version of VRChat.
I think the metaverse will do well once they can use AI to generate content, or other AI game features.
That's like adding puke frosting to a turd cake.
Yeah but have you seen what people like to eat these days?
It depends how good the AI is and how it's used etc. It would for example be neat to walk around in a massive AI generated multiplayer city that has a lot of intricate details.
Are we labeling procedural generation AI now? There have been procgen city generators for decades. Here's a particularly nice example:
https://store.steampowered.com/app/2859220/Vuntra_City/
Humans are really good at identifying the unreality of procgen because it loses the intentionality we expect from our built environments. Here's the author of that game talking about their solutions to the problem:
https://youtu.be/4MZ5-KQW3pc
Procedural generation isn't what we call "AI" now, but AI is absolutely procedural generation.
If your inputs to an LLM are prompt + seed, isn't that analogous to world config + seed in Minecraft? In fact, I would argue they both have the same problem: In an infinite world, geography has no meaning.
Why do I get bored of exploring new minecraft worlds? Because ultimately, I know the limits. I've seen cool minecraft mountains, what do I care if this time they're next to a desert but last time they were next to a lake? None of it has any meaning, there's no context. That mountain isn't a holy site for some religion, or a reminder of an ancient battle. It isn't a clever ploy by the writers of a game to set the stage for a final fight. No, it's just mountain #9057382 in minecraft world #74893124.
Midjourney's "wizard tower" with seed 1 vs seed 2? There are differences but none of them mean anything.
I played with AI image generators for a whole month and then I got bored. Not because I can't think of other prompts, but because "witch cat riding a broomstick" is something I can easily imagine in my head and that takes about the same level of effort as typing it into midjourney. And the result has the same meaning. None of my friends want to see "witch cat riding a broomstick" #A213DEF675, the novelty is gone. If they wanted to see a cat on a broomstick they could type the same prompt in.
Now, I only use image generators for D&D campaign art, and I write the campaign myself. Because I know my friends, I know what they like. I know what references they will understand, and every detail I hide in the world is something special for them.
I don't see the "metaverse" being any different. It's cool that we figured out a way to make the monkeys on the typewriters turn out stuff that is readable language instead of gibberish, but how can any work ever stand out in an infinite library?
Dead Cells uses procedural generation but does it in bespoke chunks, so while a level is different it’ll have chunks that are determined fun and then glued together by the procedural algorithm.
Yes, all of the games with the best procedural generation achieve it by heavily leveraging handcrafted pieces, but that's an indictment of procedural generation, not an endorsement.
Yes. This is the standard approach going back at least as far as the first Diablo: https://www.boristhebrave.com/2019/07/14/dungeon-generation-...
Sorta? Procedurally generated levels tend to not invite repeated playthroughs. Maybe this is a leap above previous efforts, but also maybe that isn't a problem.
I'm all for folks trying. I'm less enthusiastic it will pay off.
> Procedurally generated levels tend to not invite repeated playthroughs.
I agree with your broader point about low confidence in an AI-powered quantum leap, but in my experience this point about repeated playthroughs is actually backwards. Lots of the most popular and/or heralded games in recent years use procedural generation to enable repeated playthroughs as a core mechanic. Balatro, Hades, Slay the Spire, Diablo, Helldivers, Civilization, etc etc. Even things like drafting in MtG or board layouts in Settlers of Catan use procedural generation to increase replayability.
When the breadth of the content is just about infinite, will they care about repeated playthroughs? I think they could capture quite a few people just keen to explore and see what's around each successive corner.
I think they could do things like check the response to some areas and favour those to weed others out.
I remember spending time in RDR2 exploring the landscape, and finding the variety of moments that spawned (hunters, animal attacks, peril and whatever else). I think AI can get close enough to that quality of environment to satisfy people. As it was, the level designers were clearly inspired by the natural world which was not generated by human game designers.
Again, I'm not against people trying. I just have low confidence. It doesn't feel much different from many ideas I've experienced.
All said, the hubris of new attempts accomplishes a ton.
I dunno, it feels like when content tends towards infinite it also tends towards meaningless.
> When the breadth of the content is just about infinite, will they care about repeated playthroughs?
Read about the launch and the development of No Man's Land.
AI isn't the same as procedural, but anyways, I think there are a lot of interesting possibilities. Maybe a long ways off though.
Fair it is technically different. I question by how much? This is a lot like ML not being statistics. Somewhat true, but also largely not.
Check out Shadows of Doubt if you want an infinite supply of procedurally generated cities with intricate details. No multiplayer, though.
“Metaverse” is a moving target and has one goal for Meta: get everyone off the platforms that can eat their lunch (Chrome, iOS, Android) and onto a platform where Meta can start collecting 30% from everyone who wants to play.
People don't want to interact with the human-generated content, why would AI-generated content change their minds? I have yet to meet anyone who uses the metaverse, and it's not for lack of funding on Meta's part.
That’s because you aren’t 12 years old. Ryan George (a YouTuber) made a video where he checked out the metaverse and it was basically exclusively filled with children. He played metaverse worlds for around 5 hours straight and didn’t meet a single adult the whole time. It’s children using their parent’s old Meta Quest 2 headsets they bought during the pandemic because they were bored, then forgot about.
Until I locked my headset down my kid got onto the metaverse a few times without me realizing, and playing a bunch of random “experiences”.
The few times I (50s male) tried horizons universe I felt a visceral creepiness - that any adult I “met” was likely there because it wasn’t technically within 100meters of a school.
VR Chat was less locked down and felt more self patrolled and honest. (On the other hand I wandered a common “watch a movie” vrchat video room instance where Nazi propaganda played and the people in the room were bro-asting about being racist anti-semites.)
The population is mostly kids, but that doesn't imply it's popular with them.
Anything that can generate metaverse activities can do the same for flat-screens, which meta has been failing to convince people away from for many years despite billions of dollars of investment.
You can't blame them for trying though. Their cash cows aren't going to last forever. Actually, I don't understand why this company has such a high valuation. From all the FAANG, it seems to be the less diversified.
Making an attempt at a new line of business is fine, but normally you'd expect the funding to be reduced or stop if it wasn't succeeding.
They could have build quite a number of products with the money spent.
Metas ad platform is unfathomably large. Their tracking capabilities on mobile make them the ad market most apps use to serve ads. Apple tried to kill that but turns out it wasnt so easy and meta came back as a better tracker than ever.
> Apple tried to kill that but turns out it wasnt so easy and meta came back as a better tracker than ever.
Have any articles you can link about this? I'd like to learn more.
My gut says it's the hardware that holds it back more than anything else. It's all too heavy, clunky and inconvenient.
But the thing is that's probably not solvable without fundamental breakthroughs leading to sci-fi level technology.
Like physically there's just no way to make sunglasses a decent AR or VR display people would use all the time voluntarily.
I’m not a big PG stan, but gosh darn the submarine isn’t submerged here. It’s up in the air!