Zuckerberg’s Vision, Design vs. Computer Science, the future of Facebook/Meta
Why Meta is a bad idea...
Two weeks ago Mark Zuckerberg presented his vision for “the next version of the internet” that he named The Metaverse. He revealed an annual budget of $10 billion in R&D to turn his vision into reality, calling it his life’s work. He then changed the name of his company from Facebook to Meta.
Like me you might have wondered, what does this mean for our future? Zuckerberg is a mega-oligarch, one of the wealthiest humans in recorded history. His company controls our communication channels, shapes the contours of modern business and has massively consequential influence on our governments. Zuckerberg’s announcement clearly has implications for everyone’s future... What are they?
Like me you might have also wondered, what the hell is this guy talking about? In the first minutes of his big announcement speech Zuckerberg says “We’ll be able to express ourselves in new joyful, completely immersive ways, and that’s going to unlock a lot of amazing new experiences.” He calls it the “embodied internet”, and asserts that “the feeling of presence is the defining quality of the Metaverse.” The announcement video then goes on to demo a virtual reality that feels like a low-budget sci-fi television show for pre-teens. He starts his narration, “Imagine you put on your glasses or headset and you’re instantly in your home space…”
Huh? What do those words mean? He uses the word “embodied” to mean something close to the opposite of the definition I know. “Unlocking new experiences” and “expressing ourselves in immersive ways” are basically nonsense strings of words. I can’t help but recall the phrase that the late Philip Roth used to describe the rise of Trump: “ominously ridiculous”. The literal meaninglessness of Zuckerberg’s sentences would be laughable, except that given Zuckerberg’s power and track record, it comes off as terrifying.
I believe that Zuckerberg’s vision for The Metaverse represents an impoverishment of human experience. I predict that despite some early success and societal frenzy in the coming years, Meta’s new strategy will ultimately be a failure in the long-term.
To explain how I got there, I need to start by explaining two different approaches to social technology.
Design vs. Computer Science
I spent a few years working on social technology, first at a startup, then at Google. By “social technology” I mean the hardware and software that facilitates interactions between people in some way or another: phones, social networks, email and messaging apps, videochat, etc.
During those years I started to notice two distinct philosophies, two ways of thinking, that were animating the teams in the industry. One is rooted in computer science, the other in design.
The computer scientist uses the idea of encoding as the default approach to social technology. In computer science, “encoding” means turning something like music, video, or books into 1’s and 0’s so that a computer can transmit, or display, or understand it.
The computer scientist attempts to encode human interactions. This means more than just technically encoding the streams of videos or text that flow through a product. It means encoding the abstract concepts of social life themselves.
As an example consider the profile-page, a common feature in tech products. The computer scientist conceives of the profile-page as a way to encode identity: it turns a person’s identity, an abstract thing, into 1’s and 0’s that a computer can display. But of course even computer scientists know that identity is not a singular thing, we all have nuances to our identities, multiple versions that coexist and find expression differently in different contexts. To the computer scientist, however, this is just another feature to encode; the ultimate profile-page should be able to contain a jumble of attributes, each one represented in 1’s and 0’s, with encoded rules about how to display different versions in different contexts. The ultimate digital profile would be a life-like virtual representation of a person.
As another example consider the “like” button, a way to encode our reaction to some new morsel of information. This example is extreme, because the encoding is literally a 1 or a 0, “like” or “not-like”. The benefits of encoding, to the product creator, are clear in this case: likes can be tallied, processed in algorithms, made into games.
The designer, on the other hand, uses a totally different way of thinking that is based on the shape of objects. Designers are often taught a principle of Zen Buddhism called Ma, which very roughly translates as “emptiness”. Ma is the idea that the emptiness within and around an object is actually a part of the object, maybe the essential part. So a ring is defined as much by the void in the middle, the emptiness it encloses, as by the material it uses or the ornaments it has.
To the designer, great social technologies are best defined by what they omit, because real social life can only take place in open space. Each product effectively provides a shape for the empty space within it, for the unencoded parts. To the designer, the most successful tools in our life create shapes that allow us to express our humanity, in its full messiness and abstraction, more effectively.
As an example, consider email. Email is a technology that does not overly structure or prescribe the message, it does not try to encode any subtleties of the interaction between people. The only structure is address, subject, body. Imagine if an email asked me to check boxes about whether this was a happy or sad message, or asked me to rate other peoples’ emails for quality. These design choices would encode more, and might increase engagement with the product, but they present a clear tradeoff: by encoding more, they sacrifice fidelity in the human interaction. To the designer, this is a poor choice.
Each of these two philosophies has benefits and drawbacks. The computer science philosophy is most clearly manifested by Zuckerberg and Facebook. By encoding more, Facebook has been able to build powerful algorithms and game dynamics that feed addiction (and boost engagement), while also providing gigantic value to advertisers. Billions of daily users are proof that the computer scientist approach can be massively profitable.
The design approach is most clearly manifested by Apple. Apple’s communication products like iMessage, Facetime and Photos are the primary competitors to Facebook. There are no profile pages, no algorithmic feeds, and no targeted advertising. The products are defined by the empty space they leave open.
The problem we always had at Google when building social products was that the philosophy was never clear. Some teams championed the designer approach, others the computer science approach, and the result was often a muddled mix that failed to do either one very well. Google could have chosen one or the other (and still can), or defined some compelling third option, but it has not yet happened.
Meta and the impoverished human experience
To a humanist like me, there are two main problems with the computer science approach. First, it necessarily impoverishes the human experience. I do not believe we will ever encode the fullness, the messiness, or the inherent and inescapable contradictions of human social life. It is better to conceive of our jobs as providing more surface area for full communication. Interacting through encoded channels is less rich, less flexible, less human.
Second, as soon as we start encoding social life, we facilitate the task of bad actors. Turning something into a game means opening the door to cheaters. Encoding social life in a single database means concentrating social power in a dangerous way, power that can then be sold to the highest bidder or manipulated by political leaders.
Zuckerberg’s vision of The Metaverse is about encoding social life to an extreme degree. He hopes we all put on his headset or glasses to replace, or augment, our natural view of the world around us with a new digitized (encoded) world. This new digital world, the so-called Metaverse, will be populated by avatars, 3-d cartoony representations of us, a way to push the profile-page to an extreme of encoded identity.
This Metaverse will, without a doubt, be less humanly rich than the real-world alternative. We will lose something, many things, by encoding social life, and then replacing our view of the real world around us with this lower-fidelity digital alternative.
This Metaverse will also, contrary to claims about being “embodied”, further alienate us from our actual bodies. We are already suffering in modern life from an alienation from our bodies, from being treated like floating heads in our school and work, and there are major health consequences in our societies. (And spiritual consequences, but that’s a topic for another post or maybe another newsletter.)
Zuckerberg’s Metaverse will be like replacing real food with processed food made in a lab. Which is what terrifies me, because we know that processed food is cheaper and more addictive and has conquered the market. We also know that processed food is totally toxic for our health and has created a mass epidemic of obesity, heart problems and chronic disease.
The Metaverse will similarly be easier and more addictive than real social life, and similarly toxic for our social health. It will also open a Pandora’s box of new ways for bad actors to cheat, take advantage, and buy the newly concentrated social power.
So will Zuckerberg’s vision prevail, will The Metaverse conquer the market? I certainly hope not, and I don’t think so…
Prediction about Meta, part 1: eye contact and early success
There is one part of Zuckerberg’s vision that I believe will be a huge success: he wants to improve videochat to enable eye-contact. He mentions several times in his video and subsequent interviews that his teams are working on it. Meta’s headsets will project an image of someone in front of you (presumably either a real video or an avatar), track your eye movements, and somehow simulate eye contact in the conversation.
This has long been a holy grail in the technology sector. Conversations in the real-world are not just about an exchange of information, or an exchange of voices. There is an immense amount of non-verbal communication that happens, and much of that (not all) takes place via the eyes. The fact that videochat has become so widespread despite not allowing for eye-contact is surprising. It is one of the reasons that spending all day on Meet or Zoom is so exhausting.
My own intuition is that being able to talk with someone across the world and really look into each other’s eyes will be a magic experience, a “killer feature” that will make videochat even more widespread than it already is, perhaps much more.
I know that every major tech company is working on it, including Apple, Intel, IBM, Microsoft, probably Google and Amazon, and likely hundreds or thousands of startups. If Meta’s headsets are the first, or the best, at making it a reality, I expect that they will sell billions of them.
For the sake of humanity, I hope some other company has a better solution or is faster to reach the market. But if it is Facebook, the Metaverse will be off to a running start.
Prediction about Meta, part 2: videogames and long-term failure
Despite early frenzy around any new Meta product, and perhaps a first wave of mass adoption fueled by eye-contact, I am skeptical that Meta’s virtual reality will ever win the mainstream market in the long-term. Let’s call it a hopeful prediction.
My reasoning about long-term failure is unrelated to my reasoning here about why the Metaverse is bad for humanity. In fact, the dangers of encoding are probably arguments for future success: gamified engagement loops and easier social interactions without the complexity of the real world have been hugely successful for Facebook so far, so they might continue to win the day.
But the Metaverse is not an extension of Facebook, nor an invention of Zuckerberg’s imagination. Virtual reality worlds already exist and are already successful. It is a huge industry. Or perhaps I should say sub-industry, because it exists within the universe of Video Games.
Video Games generate more annual revenue than TV and movies combined. While a majority of the consumers in the industry are part of the “casual gaming” industry-segment on phones and social networks, the majority of revenue comes from the dedicated core of self-identified Gamers, those people who spend many hours per week playing videogames on their consoles or computers.
Somewhere around 10-15% of the population in Europe, North America and Asia are self-identified Gamers, which translates to hundreds of millions of people. Contrary to popular stereotypes, it includes people of all ages, not just teenagers. Affirming popular stereotypes, the population is around two-thirds male.
For Gamers, virtual worlds are the primary kind of game they play. The R&D budget of videogame companies actually dwarfs Zuckerberg’s newly announced budget, which has allowed them to create graphically rich, creative, extremely engaging virtual worlds.
Zuckerberg’s bet is that he can transform this existing industry into something mainstream, to expand the market by 500%. How? Presumably thanks to his new headsets that make VR feel more immersive and thanks to the social-pressure of Facebook. If all of your friends are there, you’ll want to be there, too, right?
This sounds like classic hubris to me. It is notoriously difficult to transform consumer habits. If only 10-15% of the population is attracted to virtual worlds today, why would that number change? Perhaps the number is 10-15% because that’s the natural share of people who like virtual worlds, and 85-90%, inversely, is the natural share of people who do not. People are diverse, we all have different personalities, hobbies, and tastes. Virtual worlds appeal to some and not others.
Perhaps 10-15% of people find satisfaction in solving the logical puzzles that make up games, or in controlling avatars that can restyle our real world appearances. If Gaming is a sport, as claimed by many Gamers, then maybe 10-15% is the share of the population that likes that sport instead of others, just like some minority percentage of the population likes basketball or soccer.
If Zuckerberg believes that today, 85-90% of the population do not play with virtual worlds because they don’t yet have the right headset, then he is underestimating, or ignoring, the many other more natural explanations.
Many in the tech industry point to the phenomenal growth of Roblox, a new virtual-world startup that is massively popular with teens. Doesn’t that prove, they say, that the next generation wants to live in virtual worlds? Not necessarily. Life as a teenager is unique: you are uncomfortable in your pubescent body, you have little control over your own mobility, you are eager to explore different versions of your personality and style. All of those things make virtual worlds more appealing. Most of those things are transitory, they pass after teenage years. Roblox is currently growing quickly, but as any startupper knows, current growth rates are no guarantee of future growth rates. Sometimes a phenomenon reaches a natural limit.
Perhaps Facebook’s R&D team will find the creative breakthrough that transforms virtual worlds to something much more mass-market. If their announcement video was a reveal of their best hypotheses, then I remain unconvinced that they’ve found it.
Some industry thinkers like Ben Thompson think that the functionality that will make virtual worlds mass-market are in the domain of work life and team collaboration, and he bets that Microsoft will be the winner in any Metaverse trends. I don’t see it, but the logic is at least more compelling than Facebook’s.
Of course, it is usually a bad idea to bet against Zuckerberg. He has proven his skeptics wrong time and again. For the sake of humanity, I hope this time is different.