Facebook gets one step closer to building your virtual copy

When it comes to representing yourself on social media, who you actually portray yourself as has always been a bit of a caricature. That thinking has always made it a little interesting to examine how a company like Facebook approaches avatar design for services like their VR avatar system.

Oculus Avatars have undergone a number of transformations and today they’re pushing an update that creates more robust facial expressions that are more human-like than the stiff representations that past iterations showcased. The new Sims-like “Expressive Avatars” are certainly the company’s most unsettling to date, but they’re also the most ambitious.

The company calls the update, the “culmination of user feedback and years of research and innovations in machine learning, engineering, and design.”

There’s this oft-repeated concept of the uncanny valley where things get up to a certain point of realism but then they’re just deeply unsettling because the representation is close but not quite there. That’s more than a little evident here. Oculus initially chose to veer wide when it came to structuring its avatar system based on how people actually looked, but with their latest Expressive Avatars update, things seem to be moving in a different direction.

In a blog post, the company seemed to acknowledge the risks of going all-in on realism while emphasizing that the trade-offs were worth it. They say they discovered people are simply more willing to interact with avatars when they look and behave more like humans:

Back in 2016, we made a conscious decision to avoid showing what we didn’t know in order to better represent what we knew with certainty. Since then, we’ve learned a great deal not only about how our hardware canhelpus simulate believable behaviors with higher confidence, but also about how we can use machine learning and well-understood priors to translate subtle signals into great social presence.

The new avatars boast more realistic mouth and eye movements, a small upgrade that Facebook maintains was intensely challenging to pull off.

This could be a bit of a perilous direction to choose. After all, there’s really one ideal the company can hit, recreating a perfect digital person. They’ve already copped to designing deeply human-like avatar systems; the limits here are obviously both the low power of today’s consumer systems and the inabilities of the platforms to infer very much in terms of interaction and movement aside from what’s captured precisely by sensors.

Facebook’s new Avatar system goes live today on Oculus mobile and PC platforms.

Facebook reorganizes Oculus for AR/VR’s long-haul

Facebook is again looking to whip Oculus into shape for its 10-year journey towards making virtual reality mainstream. According to two sources, Facebook reorganized its AR and VR team this week from a divisional structure focused around products to a functional structure focused around technology areas of expertise. While no one was laid off, the change could eliminate redundancies by uniting specialists so they can iterate towards long-term progress rather than being separated into groups dedicated to particular gadgets.

Facebook confirmed the reorg to TechCrunch, with a spokesperson providing this statement: “We made some changes to the AR/VR organization earlier this week. These were internal changes and won’t impact consumers or our partners in the developer community.” Oculus CTO John Carmack and Oculus co-founder/newly-promoted Head of PC VR Nate Mitchell will remain in their leadership positions within VP of AR/VR Andrew ‘Boz’ Bosworth’s hardware wing of the company.

The shift obviously communicates that Facebook believes Oculus could be running more effectively. Organizing the company around areas of expertise rather than broader divisions is probably more appropriate for a moonshot effort that can’t afford redundancies, on the other hand, keeping expertise siloed could isolate new approaches and advancements from reaching other teams. As the company builds out its first full lineup of headsets, there seems to be significant overlap in the tech problems and products bring tackled by those working on mobile and PC products.

TechCrunch reported earlier this week that the company is planning to release a new Rift headset as early as 2019, possibly called the Rift S, which will featured upgraded displays and an inside-out tracking system. The company’s “Rift 2” project, codenamed Caspar, was left behind in the reorganization, a source tells us. We can’t confirm whether any other products or concepts have been shelved.

While an immersive virtual world that users can hang out and communicate in certainly seems to fit Facebook’s broader mission, the company has spent the better part of the past few years deciding how a costly, ambitious venture like Oculus fits into its corporate structure.

First, things went smoothly. The company and its empowered co-founders were building out a developer network and prepping for the launch of their Rift headset after creating a successful partnership with Samsung for the Gear VR. Then, the company’s good fortune turned as the Rift headset was racked by expensive delays and Oculus failed to ship the company’s Touch motion controllers at launch losing some initial ground to HTC. 

By the end of 2016, it was announced that co-founder Brendan Iribe was out as CEO and that the company would be reorganizing around divisions focused on things like PC VR, mobile and content with Xiaomi exec Hugo Barra coming aboard as VP of VR to lead the new effort working directly beneath CEO Mark Zuckerberg. An additional layer of oversight has been built in since then, with Bosworth was put in charge of the company’s consumer hardware ambitions with Oculus as a central pillar. His title is now VP of AR/VR.

The absorption of Oculus deeper into Facebook’s corporate structure was a trend that soon replicated itself as the company looked to rein in the independent teams under a more cohesive vision. The culmination of this was a major executive reshuffle earlier this year that changed the landscape for how divisions within the company were managed.

Now, they’re changing things up even more.

Oculus Go

The new structure sounds like it could coordinate efforts around more general lines like hardware and software allowing insights to flow more intuitively across Facebook’s planned devices.

Given the slow adoption of VR and engineering challenges of AR headsets, which at TechCrunch’s LA conference last month Facebook’s head of AR Ficus Kirkpatrick confirmed it was building, this structure could help Oculus iterate its way to long-term success rather than just getting the next product out the door.

If Facebook is going to beat companies solely focused on AR like Magic Leap, and potential incumbent invaders like Apple if it so chooses, it needs to maximize efficiency. And if it’s going to get both developers and users excited about these next-generation computing platforms, it will have to produce products that make cutting-edge technologies feel unified and accessible. That’s a lot easier when everyone’s not stepping on each other’s virtual shoes.

Facebook’s plan to let companies it buys live independently is over

Mark Zuckerberg was quick to realize that Facebook, the largest social network in the world, doesn’t have a monopoly on all users nor can it bank on holding its position as top dog forever. Thus he instituted a policy of buying up promising rivals and integrating them into the Facebook ‘group’ in a strategy designed to be a win-win for all.

But by leaving Facebook in abrupt fashion this week, Kevin Systrom and Mike Krieger — the founders of Instagram — have shown that the social network’s vision of letting acquired businesses operate independently simply isn’t feasible.

Few large-scale acquisitions run smoothly, so it is to Facebook’s credit that Systrom and Krieger remained with the company for six years after Instagram was acquired for $1 billion in 2012.

That’s a long stretch by any tech acquisition standard, but it is still short of Facebook’s vision of entrepreneurs who continue their startup journeys inside its walls.

The Facebook Family

The original idea is a best-of-both-worlds approach: a company’s finances are infinitely secured and it can grow as needed inside the Facebook ‘family,’ with access to resources like engineering, marketing, admin, etc.

That was also the plan for WhatsApp, but founding pair Jan Koum and Brian Acton managed four and three and a half years, respectively, at Facebook following their $19 billion acquisition in 2014. VR firm Oculus, another billion-dollar purchase, lost co-founders Palmer Lucky (political scandal) and Brendan Iribe (reshuffled) three years after its deal. Ex-Xiaomi executive Hugo Barra now runs the unit as “Facebook VP of VR.”

In the normal state of tech acquisitions, getting six years — or even just three — from a founder post-acquisition is impressive. It requires a strong vision and autonomy for the incoming business.

Many founders become serial founders, even after an exit has set them up financially for life. There’s a thrill in building something new, taking sole charge and growing it. But post-acquisition, the basic dynamics change. As a founder, you call the shots — you are the boss — rather than part of the company hierarchy post-deal. Going from employee to founder requires adjustment, but back the other direction is often trickier — particularly when your business is part of a strategy for a larger one that’s rife with politics.

Facebook tried to mitigate that by promising autonomy for its founders.

Instagram’s duo retained a lot of control — Systrom was the face of the company and he reportedly approved all ads personally — while the same was true of WhatsApp, with Koum made a member of the Facebook board. Indeed, the WhatsApp founders dubbed the acquisition a “partnership” such was their insistence that things wouldn’t change under Facebook’s ownership.

Jan Koum, once a member of Facebook’s board, is said to have clashed with the social network’s management over its intentions for his WhatsApp service

Independence Vs Facebook’s Interests

But it didn’t work.

All four founders of WhatsApp and Instagram have left as Facebook inevitably sought further control of their companies in order to advance its wider goals as a business

WhatsApp, for example, has embraced business-consumer communication, is working on payments and has an advertising tie-in with Facebook. These are all features that would have troubled founders Koum and Acton, whose previous public manifesto railed against anything that takes away from a simple user experience, and particularly advertising.

After reported friction with Facebook management, Koum left in April 2018 to “do things I enjoy outside of technology, such as collecting rare air-cooled Porsches, working on my cars and playing ultimate frisbee.” Acton, meanwhile, quietly exited the year before but then tellingly wrote a $50 million check for Signal, an encrypted chat app that rivals WhatsApp, and publicly backed the #deletefacebook campaign over privacy concerns.

Over at Instagram, a similar situation seems to have happened with Systrom and Krieger. As TechCrunch’s Josh Constine reports, sources suggested that the leadership’s “weakening independence” from Facebook was a source of frustration for them that ultimately led to their untimely exit.

Reading the short farewell note from Systrom seems to hammer that home. There’s no thank you for Mark Zuckerberg, Sheryl Sandberg or any other Facebook executive. Systrom instead stated that Krieger and he are keen to explore their “curiosity and creativity again” by building new products.

Instagram CEO Kevin Systrom continued to be the service’s public face even after its acquisition by Facebook in 2012

Facebook had a good run with its independence policy, but ultimately these four exits illustrate that founders can’t be caged and tamed. While, on the other side, a buyer is always going to want to get their pound of flesh from billion-dollar acquisitions. Facebook can bend the rules and get a lengthier service from founders than most, but you can’t defy gravity forever.

While it has lost the original founders, Facebook has also seen wild success from its purchases. Instagram went from 30 million users pre-acquisition to over one billion today, while WhatsApp has more than 1.5 billion active users up from 450 million at the time of its deal.

The important question now is whether Facebook’s in-house team managing these services can continue to scale them without the inventors in place. Beyond talent, losing that original culture is a blow. These acquired services need to remain differentiated from Facebook from a consumer perspective, otherwise the entire point of owning them — the bet that the future of social networks may not be Facebook — is moot.