After the Flat Screen
Future Technologies and the Question of Design Responsibility
This entire book has been dedicated to a small window displaying our own image in the corner of a flat screen. But the flat screen is, as of today, merely an intermediate technology. Hundreds of millions of people already use AR filters, tens of millions put on VR headsets, and companies are pouring billions into augmented or mixed reality devices intended to replace the laptop, the smartphone, and the video conferencing camera. The question we posed at the beginning of this book ("What happens to a person's brain when they watch themselves while communicating?") will not simply disappear in the coming years; it will become infinitely more complex.
Self-View in Virtual Reality
In 2025, Jeremy Bailenson (the most cited researcher in this book) and his colleagues published a review in Nature Human Behaviour covering twenty-five years of experimental research on social interaction in virtual reality—from the first clunky headsets with half-second latency to lightweight devices with built-in eye tracking [1]. Out of the multitude of findings, five have direct relevance to the topic of the self-view.
First. The sense of presence (how deeply a person feels they are "inside" the virtual environment) depends less on graphical fidelity and more on the nature of the task. For skills training or psychotherapy, high immersion is highly beneficial: a surgeon practicing an operation, a pilot practicing a landing, a patient with a phobia encountering a spider or a snake. But for routine work meetings, as the review notes, full immersion is excessive and exhausting. The brain expends so much resource processing the three-dimensional environment that fewer resources remain for the content of the discussion. This confirms a thought we have returned to throughout this book: not all communication requires video. And even less of it requires virtual reality.
Second. The physical appearance of an avatar—the virtual body controlled by the user—alters the user's behavior. This is the Proteus effect, first described by Nick Yee and Bailenson in 2007 [2]: a person given an attractive avatar behaves more confidently; a person given a tall avatar becomes more assertive in negotiations. The avatar's influence on user behavior happens unconsciously. The person doesn't make a deliberate decision: "My avatar is beautiful, therefore I will be bolder." They simply and automatically become bolder. The externally observed self-image reprograms the internal state—this is the exact same mechanism that makes us feel worse when looking at a camera-distorted reflection.
Third. VR devices collect an unprecedented volume of biometric data. Head movements, eye tracking, hand gestures, facial micro-expressions—all of this forms a unique kinematic profile that can identify a user with pinpoint accuracy and, eventually, evaluate their emotional state without their consent. The self-view in Zoom is at least visible: you know you are looking at yourself and can decide how you feel about it. Biometric harvesting in VR is invisible and hidden from the user. It is the observation of the observer, entirely without their knowledge.
Fourth. The brain systematically underestimates distances in virtual environments. A conversation partner's avatar, situated logically at a distance of "two meters" in the VR coordinate system, is perceived as standing about twenty to thirty percent closer. This violates proxemics—the intuitive regulation of interpersonal distance described by Edward Hall—and reproduces the exact effect Bailenson described for Zoom: the interlocutor's face invades your intimate zone. The difference is that in VR, this sensation is experienced with significantly greater intensity than on a flat screen.
Fifth. Social behavior in VR robustly transfers into the real world. An experience lived in a virtual environment (for example, observing the consequences of deforestation from a first-person perspective) alters real-world behavior (in this example, ecological habits) for up to several weeks. Virtual experience, including the experience of observing oneself, is not isolated within virtual reality; it bleeds into the offline world.
Proteus Effect 2.0: Self-View Without the Self
The Proteus effect creates a curious situation. On a classic video call, the self-view shows your actual face—with its wrinkles, asymmetries, and tired eyes. In a virtual environment, the analog of the self-view shows an avatar. It is still "you"—you control its movements, it mimics your facial expressions—but it is no longer strictly you.
Yee and Bailenson's research demonstrated that the link between an avatar's appearance and its owner's behavior forms rapidly and operates below the threshold of awareness [2]. As mentioned, participants controlling more attractive avatars stood closer to virtual strangers and disclosed more personal information. Participants with tall avatars were more aggressive in negotiation tasks. Furthermore, the behavior persisted even after exiting the virtual environment: a person who had been "tall" in VR continued to act more confidently in subsequent face-to-face interactions.
Both tech giants and small startups are already testing AI avatars for virtual meetings—digital twins that look "like you, only better." Smooth skin, symmetrical features, professional attire, smoothed-out micro-expressions. If the Proteus effect holds true—and it does—this enhanced avatar won't just replace your image for others; it will change you. The person will begin to behave in alignment with how their digital twin looks. This requires no effort; it happens unconsciously and automatically.
From the perspective of everything we know about SVF so far, a paradox arises here. On one hand, an avatar could alleviate certain anxieties: there is no need to worry about wrinkles, blemishes, or barrel distortion. The Controller, for example, will have nothing left to control—the algorithm handles it for them. On the other hand, a new gap opens: the chasm between how the avatar looks and how the person feels inside their actual physical body. The Objectified will receive a new, impossible standard that their real face can never meet. The Hider will discover that hiding behind an avatar is even easier than hiding behind a self-view window. A self-view without a real "self" risks becoming not the solution to the digital mirror problem, but its next, more severe iteration.
Real People as Ghosts
In 2024, Michael Santoso and Bailenson investigated "video passthrough"—a technology where the user of a mixed-reality headset sees the surrounding physical world through the device's built-in cameras [3]. This is how the Apple Vision Pro and similar devices operate: you put on the headset and seemingly see the exact same room—the desk, the chairs, the window, the people. But you are not seeing it directly; you are seeing it mediately, through internal screens that create the illusion of transparency.
As it turned out, passthrough induced what the authors termed "cybersickness"—nausea and disorientation linked to a mismatch between the vestibular system and the visual signal. The technology distorts the perception of distance: objects appear closer or further away than they actually are. As a result, the body schema is disrupted—the internal sense of the boundaries of one's own physical "I," which usually remains below the threshold of consciousness but dictates all of our motor interactions with the world.
What is particularly fascinating is that real people physically present in the same room were perceived as "less present"! A person wearing a mixed-reality headset looks at a colleague sitting a meter away, and that colleague seems slightly less real. After all, they have been transformed into a digital projection on the headset's internal display. The spatial computer, which promised to erase boundaries, instead reproduces video conferencing's greatest flaw: the distancing and transformation of a live human being into a screen object. The authors termed this effect "social absence."
This is crucial for our topic for several reasons. First, mixed-reality devices are marketed as spatial computers that, unlike laptops, do not isolate the user from their surroundings. Yet, this is exactly the promise they fail to keep. Marketing materials show a person in a headset freely interacting with their family. The reality, for now, is different: the person in the headset is trapped in a translucent cocoon, viewing their family as somewhat disembodied projections.
Second, if mixed-reality technology begins to usurp the laptop as our primary work screen—and the investments of the world's largest companies point exactly in this direction—then "social absence" will move from a laboratory observation to a daily reality. A colleague sits next to you, but is perceived as a hologram. Moving the video feed from an external monitor to the inside of a display worn on the face generates new forms of the exact same dissociation discussed in Chapter 11. If on a standard video call the problem is created by your own image, in mixed reality, the presence of everyone is blurred—both yours and theirs. How quickly will we adapt to this, and what side effects will we encounter when it scales?
Filters and AI Retouching
AR filters started as entertainment: dog ears, rabbit noses, shimmering stars. But the evolution of filters quickly outgrew novelty. Today, even the default cameras on most smartphones subtly smooth skin by default—the user likely never enabled this feature and often doesn't even know it exists. Video conferencing platforms offer "Touch up my appearance" features—one checkbox in the Zoom settings, and pores vanish, skin tone evens out, and under-eye bags fade. Next-generation AI tools allow users to alter facial shape, nose size, and eye color in real-time—all framed as "cosmetic" enhancements, effectively serving as deepfakes without the intent to impersonate someone else.
As we mentioned earlier, in 2025, Sahil Mehta and Tarunpreet Narang proposed the umbrella term digitized dysmorphia, uniting three related phenomena [4]. Zoom dysmorphia: the distorted perception of one's appearance due to webcam optics—primarily barrel distortion, which widens the nose by 30%. Snapchat dysmorphia: the drive to align one's real face with a filtered image, leading people to plastic surgeons asking to "make me look like this filter." And, of course, "Instagram face": a unified aesthetic ideal featuring sharply defined contours, high cheekbones, and full lips, toward which faces of vastly different ethnic backgrounds converge after digital processing. All three dysmorphic "trends" are born of a single mechanism: the technological image supplants the internal body image and becomes the reference point for self-perception.
Mehta and Narang highlighted a phenomenon they termed professional aesthetic drift [4]. Dermatologists and plastic surgeons, interacting daily with filtered images of their patients on social media, gradually lose their objective baseline for "normal." An unedited, unfiltered face begins to seem insufficiently smooth or symmetrical to them. The standard of normalcy shifts—not just for the patients, but for the very professionals who were supposed to serve as the ultimate arbiters of objective evaluation.
Here is why this matters for our topic. In Chapter 6, discussing the Objectified, we described two layers of disconnect: one between the internal body image and what a regular mirror shows, and a second between the mirror and what a camera shows (wide-angle lens, distortion, flat lighting). The filter adds a third layer: the disconnect between the "unfiltered" image and the retouched version. A person gets used to their enhanced digital twin, and a regular reflection—in a bathroom mirror, for example—begins to cause discomfort. It's almost out of a fairy tale: the mirror that refuses to flatter becomes unpleasant to look at. The vicious cycle of self-monitoring described in Chapter 3 gets a new spin: now, the person compares themselves not just to an imagined ideal, but to a specific, retouched image of themselves they saw on a screen yesterday.
Audio as the Forgotten Alternative
Against this backdrop, it is worth remembering a format that is not merely far from obsolete, but perhaps unjustly marginalized. We have twice mentioned the Carnegie Mellon University study. They discovered that groups working in an audio-only format—without video—demonstrated higher collective intelligence than groups with their cameras on. Collective intelligence is not the sum of the participants' individual IQs; it is the group's ability to solve diverse problems better than any of its individual members could alone. And this metric was higher when the video was off.
To the reader of this book, the mechanism is already clear. Without a camera, participants do not squander cognitive resources on managing their own appearance. Turning off the video restored prosodic synchrony: participants began to more accurately read the intonation, rhythm, and pauses in each other's voices. The attention freed from the third channel was redirected to the other two: the content and the speaker's voice.
This is by no means an argument against video conferencing as a whole. There are situations where seeing the other person is critical: complex negotiations, first introductions, therapy sessions, emotionally charged conversations. But it is a necessary reminder that video is not the only—and not always the most appropriate—channel. For many work tasks—brainstorming, routine syncs, status updates—a phone call or an audio conference might be not only sufficient, but vastly more effective. The third channel, the entire subject of this book, vanishes completely the second the camera is turned off.
The Question of Responsibility
Technologies are growing more complex. Filters, avatars, headsets—every new layer generates new questions about how digital representation affects user self-perception. But the original question—the one this book started with—remains unresolved: who is responsible for the fact that the self-view is enabled by default?
In Chapter 12, we outlined three levels of responsibility: the user, the organization, and the platform. For next-generation technologies, this framework holds true—and becomes even more critical, because in VR and mixed reality, "default settings" penetrate even deeper layers of human perception.
The user can, in theory, hide their self-view, turn off their filter, or choose audio over video. But to do this, they must know the problem exists, understand its mechanism, and possess the skill and motivation to act. The book you are reading serves exactly this purpose—but not everyone will read this book.
The organization can establish cultural norms: cameras are optional, audio formats for routine meetings, a daily cap on video hours. But organizational changes are slow, and cultural norms are fragile. It takes exactly one manager convinced that "cameras must be on, otherwise how do I know people are working?" for the entire policy to crumble.
The platform can change its default settings—and this could be the most powerful lever of all. A decision made by a team of a few hundred engineers, marketers, and product managers instantly impacts hundreds of millions of users. Turning the self-view on by default was a design, interface decision. Turning it off is one, too. Adding a timer that automatically hides the self-view thirty seconds into a call requires a few lines of code. Building a soft prompt into a VR headset—"You have been observing your own avatar for twenty minutes, consider switching views"—is entirely feasible. Disabling the default alteration of facial proportions by beauty filters. Softening the effects of optical distortion via software correction.
This is not a technological problem; it is an ethical one. The architecture of choice within these platforms is currently built without any regard for psychological consequences. And until interface designers take responsibility for cognitive ergonomics, that burden falls entirely on the user. It seems logical that if society learned to demand information about the side effects of food additives, it should develop a similar demand for awareness regarding the cognitive cost of digital products.
The Mirror That Heals
The story of the digital mirror would be incomplete without acknowledging its therapeutic potential. Let us return to a story mentioned earlier in the book. In the mid-1990s, neuroscientist V.S. Ramachandran encountered a paradox: patients with amputated arms complained of agonizing pain in a limb that no longer existed. The phantom hand felt clenched into a tight fist, nails digging into the palm—and unclenching it was impossible because there was nothing to unclench. Ramachandran placed a mirror in front of the patient so that the reflection of their intact arm created the illusion of two healthy arms. The patient unclenched their intact hand, and the brain "saw" the phantom hand unclench as well. The pain subsided. Ramachandran's mirror box therapy works, and it is still used today—for phantom limb pain, stroke rehabilitation, and complex regional pain syndrome.
This example is vital for our topic because it proves once again: a mirror is neither "good" nor "bad"; it is neutral. Its effect is dictated entirely by its context. Ramachandran used the mirror consciously, purposefully, in controlled conditions, with a precise understanding of the neurobiological mechanism. He knew exactly what signal the brain would receive, and it was exactly the signal that was needed.
The self-view on a video call is also a mirror. But it affects us automatically, without awareness, without purpose, without control, and without our understanding of its mechanism. It is a digital mirror conceived as a technical feature and left running for billions of human-hours on end. As long as it remains a default interface setting rather than a conscious user choice, the digital mirror will continue to work against us. Yes, we can advocate for solutions at the organizational level and wait for initiatives from the platforms. But your best move is to hide the self-view on your very next video call.
References
[1] Bailenson, J. N., et al. (2025). Twenty-five years of social interaction in virtual reality. Nature Human Behaviour.
[2] Yee, N., & Bailenson, J. N. (2007). The Proteus effect: The effect of transformed self-representation on behavior. Human Communication Research, 33(3), 271–290.
[3] Santoso, M., & Bailenson, J. N. (2024). Video passthrough, cybersickness, and social presence in mixed reality. Study conducted at the Stanford Virtual Human Interaction Lab.
[4] Mehta, S., & Narang, T. (2025). Digitized Dysmorphia. Journal of the American Academy of Dermatology (JAAD).