In 17 years, Facebook has grown from a college social network to a potential gatekeeper for the world’s newest computing platforms: augmented and virtual reality. The Facebook Reality Labs (FRL) department sells the portal videophone and the Oculus Quest 2 VR headset upcoming line of Ray-Ban smart glasses, with more advanced AR hardware in development.
Facebook also builds and funds VR software, sometimes in Competition with smaller developers. Last year it launched its own virtual social network called Horizon in beta. It also experiments with a VR workspace system called “Infinite Office”, which can mix the real and virtual world. Especially during the coronavirus pandemic, these efforts could attract more people to work and socialize through VR.
At the same time, Facebook controls a data protection and moderation crisis. Its platform has been criticized many times Bringing online extremists together and allow discriminatory targeted advertising or harmful misinformation. These issues will almost certainly follow the business in VR and AR, complicating already sensitive issues of privacy and autonomy in these new spaces.
Andrew “Boz” Bosworth, Head of Facebook Reality Labs, described 2020 as an “enormous” year for VR in a blog post last weekHe planned to focus more on AR and Horizon in the coming year. I spoke with Bosworth through Zoom about how FRL will address today’s futuristic technology problems.
There is a pattern of social spaces being launched for VR and then realizing that there is a really large audience outside of headsets and launching on desktop or mobile. Do you see this path as feasible for Horizon?
Yes, that’s definitely a possibility that we were considering. If you want to create a social product, you want to reach people wherever they are and you need a headset that is not free when they likely already have another connected device, which will only keep some people out of access. And that doesn’t feel good to either of us.
We’re going to start in VR because if you don’t get this core mechanic properly, the rest doesn’t matter. There is already a ton of great software out there that works together to solve the task of the 2D to 2D feeling. We’re doing one of them right now. So we really wanted to have a strong foundation for VR. One of the things we’ve always talked about is how to do this across platforms, how to make this something that people can use and participate in at whatever level they can.
Facebook has been talking about a system that will allow people to put apps on Oculus Quest less exclusive than the Oculus store. What is the status of this?
I am very excited about this direction. And the status is that it comes a lot sooner than people think.
One of Pain points for Oculus the switch was to last year Facebook accounts required use the headsets. When VR becomes something you use for work, a Facebook login doesn’t necessarily seem like the best way to access it.
I think one thing that is a piece of the puzzle that you can put together with it is now that we’ve talked about linking Oculus and Facebook accounts, Facebook overall has talked more about account management. One of the product focal points for us is making it easier to manage all accounts. So Facebook Workplace accounts are a good example. It’s one of those things that I think when we look at Infinite Office we want to endorse this as one of the techniques that people can use to feel like, “Yes, I can have my Facebook account at work, and that’s what I do. ”I’ll use it in certain contexts. “
And this is in line with our general approach. We want people to have total control over themselves, right? If you want to be Batman in VR, you can definitely be Batman. We just want them to be Bruce Wayne too if they want to. And so we try to see it this way – to expand the opportunity space of “Yes, you are Batman, but you can only be Batman.” much more control over yourself, your connections and how you show yourself. This is the work we will focus on as Infinite Office evolves internally.
That seems like a bit of a reverse of the way Facebook talked about having one single identity and a uniform online presence.
I think the reality for us, especially in VR, is the ability to have a lot more control over your appearance – this is something Facebook never really looked at when you were just dealing with a profile. Many of the problems with founding Facebook were particularly concerned with authenticity on the internet. You know, nobody on the internet knows you are a dog. Authenticity was a premium feature – that you really knew who that person was and could count on it.
Now we’ve actually closed a circle in which you can be embodied. Facebook has never been able to control how you showed up in a real interaction with someone. It was never something we had to control. All of a sudden you know, in VR, we are a broker of it. So we need to give you the full wealth of self-expression you would have access to – a richer set of self-expression than in the real world, in fact. So, yes, new media requires new thinking. I don’t think it’s inconsistent. I think it’s just a confirmation of what this medium is.
How is Facebook Reality Labs’ work being limited by internet connectivity issues? Something like Horizon becomes much more difficult when people don’t have fixed, fast internet access and the pandemic has obviously brought these loopholes into relief.
I think there are two parts to this. I am always impressed with what we can do on site. I think no company has done more than Facebook to downsize artificial intelligence and run it locally on the device – for example, portal that does all face recognition and camera direction locally on the device. And that’s a great opportunity that allows these devices to be useful even when they are not interconnected in many contexts.
The second is surely, you are right. The really rich stuff we envision for Horizon may require a robust internet connection. Of course, I hope that governments around the world and not just private companies realize how much internet connectivity and access to information is increasingly a human right that we must support. But even if you only have limited internet access, we’re seeing great happiness with artificial intelligence that improves people’s experience.
Avatars require far fewer bits to express the richness of facial expression than do [Zoom] does. Law? This is a very high bandwidth connection. We can actually cut that down to a smaller number of pixels that animate my face and send those off. And you still may not have a 100 percent accurate understanding of my expressions, but a 95 percent accurate understanding and dramatically lower bandwidth costs. So there are technologies here that can actually benefit from them. Switching to avatars could even help us connect over connections with limited or low bandwidth.
You talked about how Translate privacy and security concerns on the work of Facebook Reality Labs. Specifically, what are you doing to ensure that many of the problems that have appeared on Facebook over the last year and are related to moderation do not appear in something like Horizon?
I really want to separate the moderation of content from data protection, as the topics are very different. The moderation of content is a topic that will occupy us forever. It was with us. There has always been an argument about who should be editors and who should be censors. That’s a human problem that came up, you know, as soon as the press did it. Probably before.
In terms of privacy, I feel happy. I feel that at the current height of the data protection debate we are creating a new set of media, not just in this country but in the world. We can have these conversations openly. We are in a golden age for experts in data protection and the tradeoffs that come with it. We are trying to take advantage of all the conversations that are already taking place and bring these use cases out into the world for people to discuss.
What are the business models after that? As long as Facebook Reality Labs AR and VR use the Facebook advertising model, there will be compromises in data protection at some point.
You know, in a technology tradition, we don’t really focus on the business model. They assume that if you build a great and useful thing you will find a way to make money from it.
I believe in [targeted] Advertising. I think it makes the experience people in the world have a lot better than non-targeted advertising. I think it’s hugely important for small businesses. I think it is extremely important to maximize the use of human capital. That is a debate that is a distant debate for Augmented Reality and Virtual Reality. It’s not a short-term debate.
And so I have the great luxury of not worrying about it. I have enough real problems right in front of me to address before I even think about the business model. And so we have to build it before I can think too much about it. And I am confident that in this case there will be many possibilities.
For a more short-term problem, people compared Horizon to Facebook groups. What happens when a group like QAnon starts organizing on Horizon? How to find them and decide what to do with them while threading the needle so it doesn’t look like you are scary and constantly monitor everyone?
I don’t think there is an answer to that, and there is certainly no answer that will satisfy all parties. We know that from our experience on Facebook. So I think that when it comes to content moderation, we can rely heavily on Facebook, which has suspended discussions with governments and experts and is constantly revising its guidelines as the facts change on the ground.
I don’t think you will ever expect to be in an attitude because once you do an attitude, bad actors will find little gaps. It will never end. There won’t be a single solution. I think people should expect it to be like any digital space, and frankly like any physical space that goes back in history. You need to evolve the rules as you observe the behavior.