Seems to be about VR tech so far. It's mostly about social media stuff thus far, with complaints about slow login times and how the Horizon Worlds doesn't integrate with the rest of the system and has interface issues. John wants more screen sharing for people in VR chat like environments.
John is obsessive about eliminating lag everywhere. Ironic that his facial animations are slightly lagging the audio track.
It's sort of pointless if you have to just watch it in a video like I assume we all are. I think it would be better if you were in there in the room with John (using a VR headset). Personally I'm listening to it in the background from another tab.
Yea, I actually found the avatar in the video was worse than just audio. It was just enough motion to distract my eye, while being pretty valueless in terms of communicating anything.
I ended up listening in the background, but opened up a new blank tab to hide the motion.
He seems to do an annual "update" video from wherever he's at, about minutiae of whatever he's working on at the time. It used to be QuakeCon keynotes[1] when he was at id.
I'm sure some PM somewhere in Meta is proud it's using that horrid avatar instead of an actual video. But ooft, it looks so goofy on what seems to be a serious conversation.
Their whole metaverse bet seems weirder and weirder by the day.
I "used"(found excuses not to use) this at a previous job at a mid size software company.
It was literally just a private instance of facebook. 95% of the posts were basically fluff posts from HR. No content of any substance. So basically linkedin, but less useful, because everyone on there was in the same company and so there was no real networking effect to it either.
I didn't stay at that company very long, for a multitude of reasons.
I completely agree, this absolutely horrendous for a visual presentation. A wii like character, no proper lighting, floating with no legs and the subtle fidget moving of the limbs, lack of facial expression and flapping mouth that hardly mimics the sound. What did facebook spend all their money on?
I'd rather see video or even a still picture of Carmack than this monstrosity.
VRChat has had a Quest app for a while. It has limits on avatar/world complexity because of the limited processing power, but is otherwise an identical experience, including plenty of cross-platform avatars and worlds.
I happen to have a time machine. "After Zuckerberg retired to run for prime minister of the newly united Irish state, we conducted some serious soul searching. Turns out the only CEO in the valley with anything approaching taste was Steve Jobs! Gosh, who could tell at the time? Certainly not us."
And the camera pans around, which John said is driven by someone on the production team... but, there is nothing to see there. I guess the motion is introduced to keep us from falling asleep?
But it's like Putin bombing the people he needs to convince to join him while grinding the youth out of those who had already joined: how frustrating it must be for both people's advisers to tell them to stop throwing good money after bad and frigging pivot away before all that remain are ashes and a footnote in some biography.
Zuckerberg is surrounded by yes men/women. That's the only explanation I can think of. I watched a few of their videos. The video with him talking about Zoom integration with Meta and splitting off in groups with avatars, etc. Does anyone really want to work like this?? The entire benefit of remote work is that I don't have to deal with this BS. Please stop trying to force these weird extroverted fetishes like avatar huddle groups and awkward virtual whiteboarding sessions onto remote work. We have 2d whiteboarding apps that already work!
The overall bland soulless corporateness of it all is creeping me the fuck out. These presenters are all 100% dead inside.
Skipping the commute, longer in bed, home cooking, and not having to share a toilet with the phantom sprayer - there are other benefits to remote work.
I watched a bit of the conference and they may start to pivot. They were talking about how whatever it is they were building is going to have browser integration.
It does look goofy, doesn't it? Like an elaborate joke. I'm not trying to shit on people's hard work, but it all looks so, I don't know.. bland and not interesting AT ALL. Like Second life meets that playstation thing, but even worse somehow.
Sometimes good work needs to be shit on in a constructive way.
This is bloody awful quality in visual quality and marketing. You want to show something that works well. This... this does not "work", let alone well.
I'll happily spend an evening watching vrchat shenanigans or 2.5D vtubers but this thing just goes so far beyond 'corporate inoffensiveness squeezed the soul out of it'. It's like the developers actively disdain their own work.
It seems more likely to me that they have art directors, they're just the same ones that design the weird corporate advertising art with the tiny heads. It's very "safe" for lack of a better word.
Safe works when you're selling productivity software, but not when the product is (partially) the art/design/environment.
That sounds truly painful for all the creatives involved. They must be either sad or disengaged. This is obviously a whole different ball of wax than making a copy of TikTok or whatever else they decide to clone, so you just cannot set priorities or judge quality the same way.
It’s sad on a creative level but from a career perspective, many of my friends (from when I worked in film) are earning multiples of what they earned previously, and are doing less work at the same time.
So it’s easy to just give in and do the lowered task at hand rather than fight for the high end work they were hired to do.
Part of the problem would be that you can’t have video of a person’s face while they are wearing a headset as far as I know. I expect that to remain a technical blocker even as the graphics horsepower of the headsets increase over time.
This level of quality is fine for low-end headsets like Quest. All game engines should be able to increase rendering quality on high-end systems. It's a mystery why they don't do this.
One of Meta’s moto is “we’re only 1% there” or something to that effect. I could imagine that it’d be really cool if they did that every year and it gets better and better and then we look back at the first ones and see the huge progression. I’m excited personally
Did you listen to Carmack? It's addressed immediately. He's specifically making a point about what is currently feasible and mentions how he wanted it this way since the last Connect (and said as much last time).
To be fair, a video of a VR event is way different than being in VR for that event. A comparison is a video of people watching a movie versus actually watching that movie, except more extreme. VR is far more visceral, at least when done well. A video review of, say Half-Life Alyx, is no comparison to actually being in it.
Now, that said, Meta's seems to be floundering a lot for its actual art style, there is no doubt about that!
I hope a video of the event will be uploaded to YouTube, for some reason it didn't want to play on my MS Edge Android... Only works on Chrome when on Desktop mode, but that's really not usable...
Why… why is he bottomless? (He’s represented by a Wii-like avatar. When camera angles change from close up to medium shot he’s just a floating torso in a living room set)
They did. They announced and demoed the new avatars in the Keynote, and explained why the current avatars are legless. But as with all things new, everyone is just too busy shitting on their own misunderstandings, rather than actually paying attention. :P
Your point is very well taken and I am mildly ashamed. In my own defense I mentioned that it was the one talk by him I have been unable to concentrate on due to this very problem. but thank you for doing my homework for me.
If they have to explain it maybe it wasn't a good idea in the first place? I hate these "didn't read the article and felt the need to comment anyway" just as much as the next person. But this is a presentation and they have to assume people may skip to the middle, or have sound muted, or whatever. I honestly don't care, but I'm 100% team "I will ask why there are no legs and it's not on me to do my research on this".
This isn't too far from the truth. Carmack is known for giving presentations lying on his back with his legs up in the air. this simply wouldnt work on an avatar with legs. Kudos to Meta for recognizing this before giving an embarrassing presentation.
That reasoning aside, there are an infinite number of ways to solve this that would be less "weird" to look at than simply nothing. Even a simple colored half sphere would be better than just nothing.
Its not exactly IK compute. It has more to do with it looking decent than the pure compute. There's no shoulder hip or leg tracking so its really not easy to figure out what the legs should be doing.
I'm convinced that as long as "metaverse" is a thing that involves putting on a nerdy pair of goggles to participate the whole concept is going precisely nowhere with the general population. Its like 3d TV: no one has any interest in putting special hardware on their face to participate in things like that.
The thing that bewilders me is that this has already been done with stuff like VRChat, but FB seems to think they can come along with a weird sanitized corporate version of the same thing that doesn't do anything at all new and actually attract users?
They're probably right. No one else is providing a headset as good as the quest 2 at the same price point, because no one else has $10 billion a year to burn,
More than just that, to really have a sense of your own avatar in VRChat you need a model which you either make yourself or you pay someone else to make.
Meta is going for a frictionless Memoji like avatar that let's you get going with something that can represent you.
I don't know why we should assume that a company will be successful in one area just because they've been successful in a totally different area - I wouldn't expect a text/photo based social media company to succeed at creating... whatever this is any more than I would expect Toyota to win the fast food market (although VW does make sausages...).
My prediction: the successful "metaverses" will either be created by totally new companies/indie devs, or by an established game developer (I know this stuff has some overlap with the social media wheelhouse, but no more than say World of Warcraft does).
Ideally someone creates some common open protocol for 3d/VR/AR social spaces that catches on, but I won't hold my breath. (I know there's a few projects like this around already)
Haven't they already been successful with the VR hardware? They've also had a games platform for many years now. "Totally different area" is a bit of a stretch.
I'm probably in the minority, but I won't run VRChat because of the EAC kernel-mode spyware they're using. The other competitors, Neos and ChilloutVR, are doing a poor job of even getting the basics executed. Regardless of the outcome, I'd welcome a competing platform that is run by actual professionals (unlike ChilloutVR).
So we have the Carmack fans rushing in to hear his masters voice as a floating virtual avatar in the metaverse. But I thought they weren't interested in the whole thing because of the 'Meta is dying' narrative? [0]
So far it looks like despite the so-called 'chaos' reported by the media about Meta, he is still bullish on both AR and VR at the company.
> AR is where the future is. VR is a major stepping stone.
I think this is totally wrong. VR is going to be where almost everyone spends almost all their time, and it's not even that many years away.
There was a time when humans lived in the wild, and nobody knew any different, and any talk about living in cities would have been met with derision.
Right now humans live in "reality", but in the not-too-distant future (maybe a couple of generations), people will primarily live in totally virtual worlds. There are many advantages:
- you can have as much land as you like, and you don't have to pay for it
- you can get whatever furniture or whatever home decoration you like, and you don't have to pay for it (at least the physical materials - you would probably have to pay a talented artist if you wanted to motivate them to create something bespoke; but you could probably settle for Stable Diffusion anyway)
- you can travel to visit friends instantly, no matter where they are in the world, and you don't have to deal with passports, visas, airport security, etc.
A really good VR world solves basically every pain point of the physical world.
I think people push back on this idea because they think I'm talking about VR as it appears today. I'm not. I'm talking about a version of VR where it literally feels real. Where there is no tangible difference in experience between the many versions of "reality" you can experience inside (for sake of argument...) your headset, and the singular "reality" that you experience outside your front door.
There's also the point that although HN readers may have nice large houses and comfortable lives, the average person in the world doesn't have as much space as they'd like, doesn't have as nice furniture as they'd like, has more noise from neighbours than they'd like, and so on. A really good VR solves all of these problems, and saves money at the same time. People would be crazy not to adopt it.
We'll still rely on the physical world for nutrition, of course. The same way we still rely on nature to provide our food even though we live in cities and work in offices instead of living in the wild and foraging for berries. But the proportion of human activity that will be focused on "reality" will drop from near 100% today to maybe 1% not too far in the future, the same way the proportion of activity that was focused on "nature" has dropped from near 100% 10,000 years ago to maybe 1% now.
> VR is going to be where almost everyone spends almost all their time
I believe that (current, conventional) VR will always be a niche that only sees widespread use when conventional AR headsets can deliver VR experiences
>There was a time when humans lived in the wild, and nobody knew any different, and any talk about living in cities would have been met with derision.
Cities have been an integral part of the human experience pretty much immediately after a society got farming down, and switched away from the hunter gatherer model.
>it's not even that many years away.
> I'm talking about a version of VR where it literally feels real.
This is transhumanism. We are so incredibly far away from being able to do high-precision prosthetic limbs, a fundamental baseline for understanding human/animal i/o signals. The state of the art for ocular nerves is something like 32x32 4bit grayscale - and that's not even being shipped as a product! Evenfollowing a pace of technological development that exceeds conventional televisions (a much simpler technology, that doesn't require surgery to install), you can expect to have full color and high dpi visual prosthetics in the 2040s - at the very earliest.
The timeframe for delivering this a transhumanistic experience is, in my opinion, so far down the line that talking about it in anything but the abstract doesn't contribute at all to the current discussion about AR vs VR.
> A really good VR world solves basically every pain point of the physical world.
This is not a really good VR world. This is a transhumanistic future that requires decades more research. I actually personally believe we'll get there as a species, and I'll be lucky if it's towards the end of my lifetime.
What you're describing is not "virtual reality" in the conventionally agreed upon sense, but rather something called "transhumanism".
Fwiw, Carmack spends the vast majority of his week working on AI as he is a VR/AR consultant now, and has stated it’s around one day a week (he announced a new AI startup recently), so it seems he’s more bullish on AI.
While this "metaverse" thing is worse than some "crypto NFT AR" scams I do appreciate that Facebook invest into VR R&D heavily and hopefully at some point in future some other company will benefit off their investment. I guess it's great they give funding to people like John Carmack instead of spending these money on lobbying another anti-consumer garbage.
I’m seeing a lot of R&D solely focused on giving them the chance to extract rent through future silly things like buying artificially scarce houses in fbs metaverse. I’m barely seeing any “giving back” type research like AT&T created with their research in the internet.
> I’m seeing a lot of R&D solely focused on giving them the chance to extract rent through future silly things like buying artificially scarce houses in fbs metaverse
Their end-game is very much obvious, but honestly they are so bad at it... I doubt they will manage to actually build anything successful to extract any value out of it. The reason why Facebook Metaverse is so cringey is because they have no clue how to make games or anything like it. The "real metaverse" already exist in Roblox, Fortnite and few other popular games that every teenager socialize in.
At the same time thanks to Facebook any average Joe can easily get consumer-ready VR hardware for around $400. We just really dont have many companies behind VR except for Valve and they're simply dont have enough manpower to create mass-market hardware.
My point is that we must appreciate those engineers who persuaded Zuck to spend money on VR hardware and research. Yeah their attempt at "metaverse" is laughtable, but investment into hardware is priceless.
hwers's comment lacks some context, but there is some truth to it. Consider Carmack's comment:
https://www.youtube.com/watch?v=BnSUk0je6oo @33:50 onward "A closed platform doesn't deserve to be called a metaverse, or does it?"
Making money in the Metaverse is a pillar of what makes a virtual world an actual world. (Ignoring of who does the money making for now) Scarcity is an integral part of it. Skins, models, rare items. In the virtual world it all scarcity is artificial by definition. Just like with second life and the linden dollar making it to real exchanges. Second life sold worlds as some kind of virtual realtor in the past, VRChat's users sell custom created skins and rigged models today and Meta wants to be the platform facilitating that tomorrow, but on a grand scale.
I also highly applaud Meta's contribution to FOSS btw.
I don't know what metaverse is, but I feel a lot more comfortable just playing a game like ESO and hanging out with people dancing and sitting in chairs near me. /sitchair
I can't even see how long the video in my firefox, and no jumping around. I would personally fire the people responsible for this right now, bad idea and bad implementation. But seems Zuckerberg is obsessed over this failure...
In person, there's very low latency thanks to all the engineering effort, as he mentioned. I don't understand why the capture was so incredibly terrible.
He actually mentioned there is still too much voice latency in person and he wants to fix it at the firmware level. Sad part is he's said the same thing about audio latency on Android for the past 5 or 6 Connect events going back to GearVR. The company must be completely dysfunctional. They hired a whole custom OS team and abandoned the project rather than fixing low-hanging fruit.
Holy cow and he said this was a CUSTOM BUILD to maintain frame rate! So it can’t even push this level of graphics normally. For what it showed I’d expect about 600 FPS. And the arms are jittering..Wtf?
I get that, but Bone Labs is developed by a small indie studio. It's hard to believe Meta can't make a decent looking VR experience with the amount of resources they have available. Point is, the hardware is clearly capable and is not the reason why their metaverse looks bad.
If they made it like a game it seems like they probably could make the experience better. I have no knowledge of 3D modelling, but a focus on UGC seems like a big impediment to higher quality assets.
To be fair, the actual latest stuff does look much better [1], it looks like Carmack is using an older build.
I think one of the things that's being overlooked here is scale. I've made VR Maps which pushed fidelity and the problem you have is the size absolutely explodes. Meta is intent on having automagically generated unique avatars for each user, the size of which at the scales of Meta would become astronomical.
I wonder if the majority of people designing/pushing these huge systems of weird cartoon avatar interaction systems... are a tad bit on the spectrum (Which gives them superpowers in understanding tech/algorithms....) but a side effect is the uncanny valley of these weird avatars being pushed by these teams at Meta/FB. It feels more 'creepy' to me talking to a cartoon 'uncanny-valley' representation of a person rather than just seeing their actual photo/video feed.
reading this post on "Avatars, Anthropomorphism, and Autism Spectrum Disorder"...
> A 2016 study looked measured the neural responses of children with ASD and neurotypical children when shown unfamiliar human and robot faces. The test showed that the targeted responses were present for both for neurotypical children, but only for the robot faces for the kids with ASD. “Together, these studies provide some evidence that individuals with ASD may typically process anthropomorphic rather than human faces
The biggest games are Orcs, Humans and a Lot of Brown (the Riot and most of the Blizzard portfolios). Or Toon With a Lot Of Purple (Overwatch, Fortnite, many Playstation titles). Or Brown (Modern Warfare, PUBG, CS:GO, R6). Or anime. Or photoreal (Unreal engine titles excluding Valorant & Fortnite, RDR, Cyberpunk...)
When you reach 100m users, the ideas get limited. C'est la vie.
I’m more than a touch aspy myself which makes this faux-biting rebuke that sounds cool in only my head a bit meta: but nerds figured out how to sell tech stuff to normies quite some time ago.
I dunno? If you consider paying in privacy or mental health and getting webshit sun return to be "buying something", then Zuck is probably humanities most successful salesman...
I think it's much more a facet of limited performance (as mentioned in the talk itself) on these platforms rather than anything else. Performance on headsets is a challenge on pretty much all of these apps, from VRChat to Rec Room. I also think a lot of people that look just at the pictures don't realize that presence in VR adds a lot and can compensate for some loss of graphical fidelity. It's like how the first 3D games were quite blocky and a lot less refined than the best 2D games of the era but just the ability to move your character and the camera in 3D compensated for the bad graphics, to an extent.
On PC hardware. This looks like it was built from the ground up for Quest and have a ready to customize one for each user instead of a painstakingly hand crafted one created in blender.
IIUC, users have to specifically upload a mobile version for VRC on Quest to display an avatar, rather than it trying and failing by chances. It's files missing or present difference.
Hm interesting, when I'm in VRC on Quest I see low poly avatars (haven't compared with the exact experience on PCVR mostly because if I have the Quest on I'm going to use the Quest) so I figured it was some sort of graceful degradation, but guess I'm wrong.
But why the fuck do I need this when I can multi task using screens. It just feels that once they do their final state multiverse and force us inside, we ll still have multiple screens and workspace area to organize.
The population is not even massively growing so we probably will never have a space issue ...
because it cost nothing to change a screen into 10 of custom size like you are in a nasa control room? why in the world would I buy 4 monitors when I can have it virtually. This is the beginning of the death of the monitor market
> This is the beginning of the death of the monitor market.
no. no it is not.
for one thing, people are always going to need to view screens when they are not at home or at work.
screens have a VERY long way to go before you can reproduce the clarity and just sheer number of pixels available to display clear information to the eyes in VR.
even if eye tracking is used, and even if you only render where the user is looking, the virtual monitors must be legible, and VR headsets are nowhere near that kind of pixel density, yet. they're fine for rendering scenery and walls and avatars, and conveying motion, but if you want to read a screen of text on one of your five virtual monitors without leaning in and making that monitor consume much more field of view, you're going to need LCD displays for headsets which have 10x the resolution in both X and Y than are available today. maybe only 7x.
and there will always be afterbirths like me who get motion sick instantly, and for a full 24 hours minimum, in VR no matter how good the experience is.
Not everyone has space, not everyone wants to spend tons of money (generally the same people who don't have space) on 5 displays and the OS config to make 5 displays work. Some people travel a lot for work. There will always be enthusiasts like the people who spend absurd amounts of money on fancy mechanical keyboards, but once VR resolution becomes good enough for work, anyone who wants/works better with large screen real-estate or travels for work a lot will just dump monitors.
BTW if you're actually interested in this, Carmack talks about this and the state of the Oculus around ~ 56 min in the linked video.
While I agree with most of your comment (I don't have space — it's ridiculously expensive! Wouldn't mind another monitor but it's also a bit expensive…)
> absurd amounts of money on fancy mechanical keyboards
Fancy mechanical keyboards aren't really expensive? (compared to monitors, or space, or even really VR headsets…) Most of the people I know who have them also have them because RSI. That forces you into being an enthusiast whether you want it or not. But a single trip to the doctors saved and the keyboard pays for itself.
>The population is not even massively growing so we probably will never have a space issue ...
Our massive real estate issues that are pricing people out of having enough space to live in the last decades, have less to do with the population growth in the world in this time frame and more to do with immigration and economic and building policies that must see the asset prices always go up by significant margins to "beat the market".
There's also the issue of location and internal and external migration where given a constant worldwide population, everyone is migration to the handful of cities with jobs, infrastructure and QoL, pushing prices up. I've seen this in my home country that saw a massive population decline yet a massive increase in real estate prices proving that the price of housing is not determined by taking the number of houses in the country and dividing it by the number of people in the country giving us the supply/demand ratio determining the housing price factor, as not people are competing for real estate but piles of money are, since it's also a speculative investment vehicle, and so the bigger piles of money are winning, regardless of the total houses to people ratio.
I've gone a bit off topic but my point is that the struggle for affordable living space is real and is gonna get worse despite the world population not increasing that much, as in traditional capitalist fashion, the banks, companies and people who already have wealth and assets will use it as leverage to aquire more of it, pushing the prices up, at the detriment of those trying to enter the market now with less capital.
TL;DR: The housing situation has nothing to do with population growth trends but everything to do with growing piles of money competing for a pie that's relatively fixed in size. This cannot end well.
The push for low fidelity, as usual, and as mentioned in this very talk, is the limited power of these mobile systems pushing dual 2k displays at 90Hz.
But running at 120fps at basically 4k resolution while scanning 5 camera inputs on the same low power mobile CPU? I think not.
And remember these scenes can be modified by users and are loaded as needed.
This is not like a precooked scene in a computer game where designers spend weeks fine-tuning the lighting model.
I notice pretty much all quest apps use ambient lighting only which gives them this fake look. Some nice spotlights would do a world of good. But I guess anything better than this is just too expensive for the system.
You actually don’t get the “uncanny” feeling from this? Compared to non-Meta avatar systems and game graphics? Are non-gamers less susceptible to it? Is that how these avatar systems goes through the front door???
It's not the number of polygons that is the problem, its how they are arranged.
They need to hire an better Art Director.
They could take some polygons out of the glasses and give the characters legs at least!
Update: They are a actually doing a lot there. The lips and eyebrows are moving. It has skinned animation. When I started in the industry we were designing characters that would inanimate without skinning, everything was static like a wooden puppet.
I'm pretty confident this has more polygons and higher resolution textures than a gameplay scene in Final Fantasy XII. Those models had 256x256 textures and 1500 polygons total per character. Just the glasses virtual-Carmack is wearing are probably pushing 200+ polygons when they could be 3 quads and a clever texture. I was going to think maybe they wanted to avoid using transparency in their textures, but he's standing in front of a large virtual glass door.
But still I agree with the other commenters, they should have gone MORE cartoony and more simple. This is just a simultaneously creepy and goofy middle-ground.
It's coming, to, of all things, Second Life. Second Life is adding "puppeteering", and this works with head tracking, face tracking, and full body tracking. Users have been working on VR viewers. (The client is open source.)
It's all very experimental, but there is a meeting in Second Life every two weeks[1] where all of this is demoed. Some of the attendees are rigged for full body tracking, and, with good quality Second Life avatars, which have far more detail than Meta or VRChat, and can do facial expressions, they look reasonably good. It's striking how someone with a full rig dominates a meeting without even trying. They're so much more present than the people sitting at keyboards.
Hardware is a big problem. You need a good gamer PC to run all this and get the full effect.
And a VR headset. And tracking gear. This is a real problem in the era of the $99 Chromebook. It's not getting cheaper, either. The first NVidia graphics card that was good enough for this was the NVidia 1060 with 6GB VRAM, in 2016, introduced at $250. It still costs about $250 to get that level of GPU power from NVidia.
People who really want to be looked at - performers, influencers, etc - will need a full tracking rig. The audience doesn't.
I suspect its just easier and faster (in terms of cpu and network throughput) to translate human facial expessions to a digital represebtation and apply it to lofi animated character than using something more hi def.
I initially thought the same, but using even the current gen avatars for multiplayer on the Quest 2 feels great and is surprisingly convincing. The speech to mouth movement detection looks natural and combined with the body language being accurately represented with reverse kinematics I felt like I was colocating very realistically.
Is there any good way for me to try this out without buying a $1500 piece of hardware? My instinct is that Oculus' VR vision doesn't make any sense, but I want to try it out before being confident that my instinct is correct.
Thanks to FB you can buy $400 "Meta Quest 2" and experience like 95% of good VR content. Yeah it somewhat worse than Valve Index for VR enthusiasts, but it also much cheaper.
Every single one of my childhood heroes has committed suicide before my eyes by doing stuff like this. I always considered them having some solid morals and being wealthy enough would make this improbable but boy I was wrong. Oh well, they can't kill the old memories, yet.
To be honest Carmack have a good reasons to work on it even if it means work for TheEvilCorp: first of all if you listen to him he is truly interested in advancing VR tech. So he doesn't do it just for the money.
Also I guess he run out of money when Armadillo Aerospace has failed and at least before he talked like he want to make second attempt in space industry and this will require a lot of money.
Good for him then, but AFAIK he is still very much into space and I hope he'll have another chance even if it's gonna be funded by money Facebook gave him.
PS: Might be something changed by now, but I hope not. I really need to check latest podcasts / interviews with him.
Yes, it's very unfortunate that he is associated with Facebook now :| Such great technology, that I'll never use until other companies are able to catch up.
The idea of using any sort of VR tech that Facebook controls literally disgusts me. Can't even imagine the amount of information they will control on people with such things...
The Vive is just as good as the Oculus (VR games with a decent gaming PC hooked up is a much better experience than the all in one stuff that's running on mobile hardware).
I'd also look out for the PSVR 2 when that releases - just as walled garden as Oculus but it's not Facebook.
290 comments
[ 3.6 ms ] story [ 479 ms ] threadI think it's live now and hopefully there'll be a transcript later.
John is obsessive about eliminating lag everywhere. Ironic that his facial animations are slightly lagging the audio track.
I ended up listening in the background, but opened up a new blank tab to hide the motion.
[1] https://www.youtube.com/watch?v=4UBmmpt5iJ8
[1]: https://archive.org/details/john-carmack-quakecon-keynotes
Their whole metaverse bet seems weirder and weirder by the day.
Then it's will turn into the one true cringe generator.
“Please help me celebrate my Metaversary…”
“Nobody truly appreciates how hard us weebles work to get our Zuckabadges. Next time you see a weeble, give it a ‘/clap’!”
Oh sweet lord this will be horrific.
Also, your colleagues are your “Metamates”. I’m not even kidding.
It was literally just a private instance of facebook. 95% of the posts were basically fluff posts from HR. No content of any substance. So basically linkedin, but less useful, because everyone on there was in the same company and so there was no real networking effect to it either.
I didn't stay at that company very long, for a multitude of reasons.
Anything the VR stuff is doing feels like a "good on paper" awful distraction.
I'd rather see video or even a still picture of Carmack than this monstrosity.
This is trying to be vaguely human in hand and lip movement and it looks like bad marionette.
Hopefully it gets better but right now I’d rather just listen to the audio.
I’m hoping this is just an expensive example of the Sunk Cost Fallacy because otherwise it’s perplexing and eerie.
Facebook missed the boat on owning an OS or phone, so their only hope in protecting access to their apps is by owning a brand-new platform.
Which means that the company is all-in on VR, regardless of whether or not the result will be any good.
Also, most of us will never work on anything that anyone will ever give a shit about.
Even if this particular project fails, this is mildly more interesting than yet another CRUD app.
The overall bland soulless corporateness of it all is creeping me the fuck out. These presenters are all 100% dead inside.
https://makeavideo.studio
This is bloody awful quality in visual quality and marketing. You want to show something that works well. This... this does not "work", let alone well.
From the evidence, Meta doesn’t have any.
Safe works when you're selling productivity software, but not when the product is (partially) the art/design/environment.
<shiver>
However talking to the folks there, it’s very much an engineering project manager led environment, and there’s just no value on good art direction.
You can have the best of the best but you need to put them to good use, and Meta clearly isn’t.
So it’s easy to just give in and do the lowered task at hand rather than fight for the high end work they were hired to do.
Now, that said, Meta's seems to be floundering a lot for its actual art style, there is no doubt about that!
[1]: https://youtube.com/watch?v=hvfV-iGwYX8&feature=share&si=EMS...
[2]: https://youtube.com/watch?v=hvfV-iGwYX8&feature=share&si=EMS...
The headset he’s wearing can’t see his legs reliably. It’s just the physics of what can be seen from there.
Add to that, inverse kinematics to lock the legs to the floor is expensive to compute (when you’re dealing with such a small frame budget)
They did announce legs support in the keynote though.
Meta is going for a frictionless Memoji like avatar that let's you get going with something that can represent you.
Edit: My mistake — Zuck added a poke button.
My prediction: the successful "metaverses" will either be created by totally new companies/indie devs, or by an established game developer (I know this stuff has some overlap with the social media wheelhouse, but no more than say World of Warcraft does).
Ideally someone creates some common open protocol for 3d/VR/AR social spaces that catches on, but I won't hold my breath. (I know there's a few projects like this around already)
So far it looks like despite the so-called 'chaos' reported by the media about Meta, he is still bullish on both AR and VR at the company.
[0] https://news.ycombinator.com/item?id=33087535
So of course he's not going to trash talk it.
> But I thought they weren't interested in the whole thing because of the 'Meta is dying' narrative? [0]
I am happy to listen to a smart person who is very clever talk about things I'm interested in, even if I disagree with them on key points.
I think this is totally wrong. VR is going to be where almost everyone spends almost all their time, and it's not even that many years away.
There was a time when humans lived in the wild, and nobody knew any different, and any talk about living in cities would have been met with derision.
Right now humans live in "reality", but in the not-too-distant future (maybe a couple of generations), people will primarily live in totally virtual worlds. There are many advantages:
- you can have as much land as you like, and you don't have to pay for it
- you can get whatever furniture or whatever home decoration you like, and you don't have to pay for it (at least the physical materials - you would probably have to pay a talented artist if you wanted to motivate them to create something bespoke; but you could probably settle for Stable Diffusion anyway)
- you can travel to visit friends instantly, no matter where they are in the world, and you don't have to deal with passports, visas, airport security, etc.
A really good VR world solves basically every pain point of the physical world.
I think people push back on this idea because they think I'm talking about VR as it appears today. I'm not. I'm talking about a version of VR where it literally feels real. Where there is no tangible difference in experience between the many versions of "reality" you can experience inside (for sake of argument...) your headset, and the singular "reality" that you experience outside your front door.
There's also the point that although HN readers may have nice large houses and comfortable lives, the average person in the world doesn't have as much space as they'd like, doesn't have as nice furniture as they'd like, has more noise from neighbours than they'd like, and so on. A really good VR solves all of these problems, and saves money at the same time. People would be crazy not to adopt it.
We'll still rely on the physical world for nutrition, of course. The same way we still rely on nature to provide our food even though we live in cities and work in offices instead of living in the wild and foraging for berries. But the proportion of human activity that will be focused on "reality" will drop from near 100% today to maybe 1% not too far in the future, the same way the proportion of activity that was focused on "nature" has dropped from near 100% 10,000 years ago to maybe 1% now.
I believe that (current, conventional) VR will always be a niche that only sees widespread use when conventional AR headsets can deliver VR experiences
>There was a time when humans lived in the wild, and nobody knew any different, and any talk about living in cities would have been met with derision.
Cities have been an integral part of the human experience pretty much immediately after a society got farming down, and switched away from the hunter gatherer model.
>it's not even that many years away.
> I'm talking about a version of VR where it literally feels real.
This is transhumanism. We are so incredibly far away from being able to do high-precision prosthetic limbs, a fundamental baseline for understanding human/animal i/o signals. The state of the art for ocular nerves is something like 32x32 4bit grayscale - and that's not even being shipped as a product! Evenfollowing a pace of technological development that exceeds conventional televisions (a much simpler technology, that doesn't require surgery to install), you can expect to have full color and high dpi visual prosthetics in the 2040s - at the very earliest.
The timeframe for delivering this a transhumanistic experience is, in my opinion, so far down the line that talking about it in anything but the abstract doesn't contribute at all to the current discussion about AR vs VR.
> A really good VR world solves basically every pain point of the physical world.
This is not a really good VR world. This is a transhumanistic future that requires decades more research. I actually personally believe we'll get there as a species, and I'll be lucky if it's towards the end of my lifetime.
What you're describing is not "virtual reality" in the conventionally agreed upon sense, but rather something called "transhumanism".
I don't see the narrative you're talking about in that link.
An example from FB please?
At the same time thanks to Facebook any average Joe can easily get consumer-ready VR hardware for around $400. We just really dont have many companies behind VR except for Valve and they're simply dont have enough manpower to create mass-market hardware.
My point is that we must appreciate those engineers who persuaded Zuck to spend money on VR hardware and research. Yeah their attempt at "metaverse" is laughtable, but investment into hardware is priceless.
For publications in general: https://research.facebook.com/publications/
Making money in the Metaverse is a pillar of what makes a virtual world an actual world. (Ignoring of who does the money making for now) Scarcity is an integral part of it. Skins, models, rare items. In the virtual world it all scarcity is artificial by definition. Just like with second life and the linden dollar making it to real exchanges. Second life sold worlds as some kind of virtual realtor in the past, VRChat's users sell custom created skins and rigged models today and Meta wants to be the platform facilitating that tomorrow, but on a grand scale.
I also highly applaud Meta's contribution to FOSS btw.
is this parody? The quality of this is really really bad, almost ps2 level bad
Using this for presentation is really unwise.
[1]: https://www.oculus.com/experiences/quest/4215734068529064/
TLDW: Horizons is optimized for UGC and being dynamic. Horizons stuff isn't built in modelling tools by professional modellers.
I think one of the things that's being overlooked here is scale. I've made VR Maps which pushed fidelity and the problem you have is the size absolutely explodes. Meta is intent on having automagically generated unique avatars for each user, the size of which at the scales of Meta would become astronomical.
https://youtube.com/watch?v=hvfV-iGwYX8&feature=share&si=EMS...
reading this post on "Avatars, Anthropomorphism, and Autism Spectrum Disorder"...
> A 2016 study looked measured the neural responses of children with ASD and neurotypical children when shown unfamiliar human and robot faces. The test showed that the targeted responses were present for both for neurotypical children, but only for the robot faces for the kids with ASD. “Together, these studies provide some evidence that individuals with ASD may typically process anthropomorphic rather than human faces
https://codebaby.com/avatars-anthropomorphism-and-autism-spe...
When you reach 100m users, the ideas get limited. C'est la vie.
VRChat avatars have personality. On hardware now 6 years old.
People are reacting to the lack of personality, not the lack of immersion.
The population is not even massively growing so we probably will never have a space issue ...
no. no it is not.
for one thing, people are always going to need to view screens when they are not at home or at work.
screens have a VERY long way to go before you can reproduce the clarity and just sheer number of pixels available to display clear information to the eyes in VR.
even if eye tracking is used, and even if you only render where the user is looking, the virtual monitors must be legible, and VR headsets are nowhere near that kind of pixel density, yet. they're fine for rendering scenery and walls and avatars, and conveying motion, but if you want to read a screen of text on one of your five virtual monitors without leaning in and making that monitor consume much more field of view, you're going to need LCD displays for headsets which have 10x the resolution in both X and Y than are available today. maybe only 7x.
and there will always be afterbirths like me who get motion sick instantly, and for a full 24 hours minimum, in VR no matter how good the experience is.
I'll keep my monitors, thanks.
BTW if you're actually interested in this, Carmack talks about this and the state of the Oculus around ~ 56 min in the linked video.
> absurd amounts of money on fancy mechanical keyboards
Fancy mechanical keyboards aren't really expensive? (compared to monitors, or space, or even really VR headsets…) Most of the people I know who have them also have them because RSI. That forces you into being an enthusiast whether you want it or not. But a single trip to the doctors saved and the keyboard pays for itself.
But they're also just better keyboards too.
I find this tendency funny. It's kinda crazy we keep building new very hard innovative software to avoid needing to fix boring software.
I bet vscode will never support multiple monitors, even in vr.
Our massive real estate issues that are pricing people out of having enough space to live in the last decades, have less to do with the population growth in the world in this time frame and more to do with immigration and economic and building policies that must see the asset prices always go up by significant margins to "beat the market".
There's also the issue of location and internal and external migration where given a constant worldwide population, everyone is migration to the handful of cities with jobs, infrastructure and QoL, pushing prices up. I've seen this in my home country that saw a massive population decline yet a massive increase in real estate prices proving that the price of housing is not determined by taking the number of houses in the country and dividing it by the number of people in the country giving us the supply/demand ratio determining the housing price factor, as not people are competing for real estate but piles of money are, since it's also a speculative investment vehicle, and so the bigger piles of money are winning, regardless of the total houses to people ratio.
I've gone a bit off topic but my point is that the struggle for affordable living space is real and is gonna get worse despite the world population not increasing that much, as in traditional capitalist fashion, the banks, companies and people who already have wealth and assets will use it as leverage to aquire more of it, pushing the prices up, at the detriment of those trying to enter the market now with less capital.
TL;DR: The housing situation has nothing to do with population growth trends but everything to do with growing piles of money competing for a pie that's relatively fixed in size. This cannot end well.
And remember these scenes can be modified by users and are loaded as needed.
This is not like a precooked scene in a computer game where designers spend weeks fine-tuning the lighting model.
I notice pretty much all quest apps use ambient lighting only which gives them this fake look. Some nice spotlights would do a world of good. But I guess anything better than this is just too expensive for the system.
They need to hire an better Art Director.
They could take some polygons out of the glasses and give the characters legs at least!
Update: They are a actually doing a lot there. The lips and eyebrows are moving. It has skinned animation. When I started in the industry we were designing characters that would inanimate without skinning, everything was static like a wooden puppet.
But still I agree with the other commenters, they should have gone MORE cartoony and more simple. This is just a simultaneously creepy and goofy middle-ground.
Imagine the opposite - photorealistic avatars. There would be no interaction! You'd spend all your time rendering.
You mean like... video?
It's coming, to, of all things, Second Life. Second Life is adding "puppeteering", and this works with head tracking, face tracking, and full body tracking. Users have been working on VR viewers. (The client is open source.) It's all very experimental, but there is a meeting in Second Life every two weeks[1] where all of this is demoed. Some of the attendees are rigged for full body tracking, and, with good quality Second Life avatars, which have far more detail than Meta or VRChat, and can do facial expressions, they look reasonably good. It's striking how someone with a full rig dominates a meeting without even trying. They're so much more present than the people sitting at keyboards.
Hardware is a big problem. You need a good gamer PC to run all this and get the full effect. And a VR headset. And tracking gear. This is a real problem in the era of the $99 Chromebook. It's not getting cheaper, either. The first NVidia graphics card that was good enough for this was the NVidia 1060 with 6GB VRAM, in 2016, introduced at $250. It still costs about $250 to get that level of GPU power from NVidia.
People who really want to be looked at - performers, influencers, etc - will need a full tracking rig. The audience doesn't.
[1] https://community.secondlife.com/blogs/entry/6509-introducin...
does it?
I think it is weird for a different reason: it looks like 1990s tech - https://en.wikipedia.org/wiki/Active_Worlds
You guys are pulling my leg, right?
Also I guess he run out of money when Armadillo Aerospace has failed and at least before he talked like he want to make second attempt in space industry and this will require a lot of money.
PS: Might be something changed by now, but I hope not. I really need to check latest podcasts / interviews with him.
The idea of using any sort of VR tech that Facebook controls literally disgusts me. Can't even imagine the amount of information they will control on people with such things...
I'd also look out for the PSVR 2 when that releases - just as walled garden as Oculus but it's not Facebook.