Thesis: The platform that efficiently allow our multiple realities to collapse is the next platform that will help us think beyond what we know today.

Arguments :

  • the web is already at the center of the majority of mediated exchange (VoIP, wiki encyclopedia, scientific research repositories, MOOC, MMORPG, social networks, etc)
  • economy of scale, network effects (Metcafe's law) and trust in remote platforms are driving web based adoptions of tools and content
  • there is an accelerating transition from local interface (e.g. Word Processor), formatted content (e.g scientific publications on arXiv) and backend (e.g. database) to the cloud
  • there is a shift, arguably a come back, from app stores (where already a significant share of content is web based) to web apps
  • technologies required for XR are becoming available on the web and will adopt those trends
  • permission-less innovation is the most efficient way to learn a new media and uncover new usages

Most of those arguments need sources.

Introduction: We are using the web hundreds of times per day, usually directly, opening our mailbox for example, but even more often unknowingly, using an "app" on our smartphone that is in fact relying entirely on web technologies (HTTP, WebRTC, HTML5, CSS). The new interfaces of augmented reality (AR) and virtual reality (VR) gaining popularity those last few years thanks to the significant price drop of sensors may appear quite far away from the web. Yet thanks to the efforts of the lead browser vendors (Google Chrome, Mozilla Firefox, Microsoft Edge) AR and VR are slowly but surely arriving on the web. In fact the role of the browser to display a simple page with solely text and links first then images to finally within the last few years 3D is constantly evolving. Moving from 3D to VR happened without much fireworks since 2016 at Mozilla hen Google with basic experimentation. Since and thanks to the collaboration of all those actors but also content producers or distributor the web could very well be shifting to become the first distribution channel of all immersive content: AR, VR, MR... to XR today.

The World-Wide-Web or more casually the web is a well defined technological construct. It is measurable in term of size, electrical requirements, number of views per time period and more. Yet to simplify we can start from the bare minimum : 2 pages with a link. This very small and simple construct is the foundation for the web as we know it today.

This deceptively simple construct comes with an important set of properties namely that :

  • each page can be stored physically anywhere as long as they are on the same network
  • the owner of each page can be different
  • there is no need to ask permission to link to another page
  • the format of the page is written in an open format allowing anybody to build and provide tools to create, modify, host a page

Those properties might seem very basic but they allow for quite unique behaviors namely :

  • the network of pages can grow without are pre-defined structure
  • there is no authority indicating which page is good, bad or even relevant
  • the set of tools to interact with the network can freely grow in complexity as long a consensus remains
  • the topology of the underlying network infrastructure can change

It is precisely those properties allowing the web to grow that made it this tool or platform to think so central to our daily lives. The underlying technologies are important and even though no perfect do work but the determining factor was the ability to grow based on the efforts of content creators and their desire to connect to other content.

This set of desire and efforts remain regardless of the media, the web started as text then adopted images, animations, videos but now is gradually adopting XR. Those new media will consequently automatically benefit from the properties of the web that allowed it to grow so large and so fast.

Growth itself is interesting but isn't necessarily associated with value or usefulness. There might be a vast amount of content with very low diversity or quality. Without taking strong assumptions we can rely on usage as a metric of usefulness, if the tool is used we can assume it is useful at least in some ways to its user. This perspective is analogous to how we perceive our environment. We do not perceive through our imperfect senses an objective description of the world around us but rather an interface of the properties of the world based on their relative usefulness to us as an active agent of this world. We do not perceive objects but affordances to task relative to a goal. We behave day to day as signal processing organisms avoiding threats and securing resources needed to maintain the homeostatis required for our survival. This perspective has the quite profound implication to transform everything around us, not just man-made tools, to interfaces for our shared or distinct goals. It also simplifies how we can think of new media like virtual reality, augmented reality, mixed reality or any other engineered reality. Those engineered realities are intermediary interfaces that, when successful, allow us to reach our goals more efficiently.

Consequently we can analyze the future of engineered realities through the filter of usefulness relative to a goal. What the web allow in this context is, like before, to allow for an efficient way to create, test, share then curate those engineered realities. What it also allows, the same way we were able to connect the first 2 pages of the web together, is to connect two engineered realities regardless of their owner. This might seem trivial but here I believe is both the distinctive advantage of web page compared to any other medium beyond that what will make the web the platform for future thoughts.

What is already currently possible is for a content creator build a virtual world, say a simulation of the social structure during medieval time in continental Europe. In itself being able to build a world and let anybody experience it instantly is incredibly powerful and, until tried once, arguably impossible to fully comprehend. What is also possible is for somebody else entirely unknown from the content creator to build another virtual world allowing to explore the social structure during the same time period in China. What is unique to the web though is the ability then for the second creator to link back to the first virtual world. This trivial action though enable entirely new insights namely the ability to compare, to be able to potential identify commonalities and differences. Beyond that what is I believe even more important is the ability to step back and try to identify a pattern namely are those commonalities and differences applicable to other object within the same group? If we create a third virtual world from another culture will it exhibit all those properties, some of them, are they predictable and if so, why? This again is what is making engineered realities on the web so unique, they are becoming manipulable objects that can now be :

  • created
  • shared but also remixed
  • linked and then indexed

The ability to conceive realities as nearly trivial manipulable object is very powerful in the sense that what is arguably the most personal to us, our wordviews, can then be more efficiently change and ideally improved. Even though as claimed before we only see the world around us as interfaces the quality of those interfaces directly impact our well being. If we have incorrect interfaces our livelihood might be at risk. Consequently being able to efficiently manipulate engineered realities is not only possible but very valuable.

The previous example showcased how multiple virtual reality experiences can be linked today but it is already possible to do more. Namely in addition of going from a virtual world to another or list them it is possible to bring content from a virtual world to another. Let's imagine that in the first example of virtual world in continental Europe during medieval time one gets attributed randomly resources that will impact it's social status one can imagine being able to bring those to the virtual world representing China at the time. Consequently it is possible to create a causal link between actions in one world and another. Those virtual worlds become de facto causally linked.

Another possibility is to link the virtual reality example to an augmented reality example. Say you are visiting a castle in the French country side after having visited the virtual world on the same topic. You can now use your AR device to see superimposed on the physical environment around you but, again based on the status you previously had based on your allocated resources, the behavior of virtual character will be adapted. If you had a very low social status most places will be banned or at least with a thorough unpleasant set of questions and verification.

Those links and their ability to make the experience continue from one virtual world one to another or to an augmented reality overlay is not trivial as it allows an isolated experience to permeate within a larger context.

How the social structure in medieval time changes one's worldview might not be obvious (even though if resources are allocated randomly the extremely high chance of being of lower status would most likely make one truly appreciate their current life situation) but one can imagine quite a few other experiences from the mundane to the rare awe evoking moments. What the web once again allow is indexing (or more casually bookmarking) which is itself a both personal and thoughtful process. By indexing engineered realities that were indeed useful to me I can create a curated selection of experience that can again be lived again in the appropriate context. This very action of curating useful engineered realities is already becoming a new skill but is obviously creating a meta-interface. Our curated list of useful engineered realities is basically becoming a proxy for our constantly evolving worldview. One could even imagine that sharing one's bookmarked engineered realities is becoming an explicit way to provide to someone else a glimpse through the set of interfaces that one is using to navigate and interact with the world.

To conclude the web isn't just yet another distribution channel for new immersive content of AR or VR, it is already the most efficient way to build test and share content. It allows and fosters permissionless innovation that are so crucial for an emerging new medium. Finally the web allows to link content and those beyond this to connect content creators and consumers in a way that no other solutions allows in XR. The web will consequently be the medium thanks to which the interconnection of XR content will let blossom a future we can barely imagine today.

Finally, the link as a primitive is what will allow us to make engineered realities manipulable objects allowing us to think beyond what we know today.

Themes to clarify :

  • interfaces (Donald Hoffman)
  • information layers (Blair MacIntyre)
  • web as social mediation (networked AFrame)
  • web as invisible yet pervasive framework (mobile apps as webview)
  • web as leading information interaction mean
  • web as a radically new way to create and consume AR/MR (rapid prototyping e.g. Glitch)
  • web as a radically new way to create and consume VR (rapid prototyping and link traversal)
  • specification (WebXR) and implementation work as complex dynamics (with W3C orchestration)

To add

  • Importance of links (cf title of the journal)
  • Importance of plurality of perspective and models (puridisciplinarity as a way to learn about the world)
  • defining what the "web" actually is
  • AR cloud
    • difficulty of (re) localisation
  • metaverse
  • VR vs AR vs XR
    • VR sharing perspective as first person point of view
    • AR sharing layers as informed interface
    • XR juggling both back an forth, eventually allowing one to modify the other

Own motivations


  • Les objects dans l'espace, la planification dans l'action, B. Conein, E. Jacopin, Raisons Pratiques 1993
  • As We May Think, Vannevar Bush
  • Bringing Network Effects to Pervasive Spaces
  • VirtualRealityInterface#References

Nous utilisons tous le web des centaines de fois par jour, souvent de facon explicite, en ouvrant notre boite mail par example, mais encore plus souvent de facon implicite, en utilisant une "app" sur notre telephone mobile qui repose entierement sur les technologies du web (HTTP, WebRTC, HTML5, CSS). Les nouvelles interfaces de realite augmentee (RA) et de realite vituelle (RV) qui redeviennent populaire ces dernieres annees en particulier grace a la reduction drastique du prix des senseurs semblent bien loin du web. Pourtant suite aux efforts des principaux navigateurs web (Google Chrome, Mozilla Firefox, Microsoft Edge) RA et RV arrivent doucement mais surement sur le web. En effet le role d'un navigateur affichant une simple page web constituee de texte et de liens premierement puis des images ensuite et depuis quelques annees de la 3D ne cesse d'evoluer. Le passage de la 3D a la realite virtuelle s'est deroule discretement depuis 2016 a Mozilla puis Google par de simples experimentations. Depuis et grace a la discussions de tous ces acteurs mais aussi de createurs de contenu et de distributeurs le web pourrait bien etre en phase de devenir le premier moyen de distribution de contenu immersif: RA, RV, RM, ... puis RX maintenant.

Puis comme themes a aborder :

  • clarifications sur la notion d'interfaces (Donald Hoffman)
  • clarifications sur la notion de couches d'information (Blair MacIntyre)
  • web comme support de mediation sociale (networked aframe)
  • web comme support technique transparent (mobile app)
  • web comme support d'interaction informationnel
  • web comme completement nouvelle facon de creer puis d'utiliser la RARM (rapid prototyping)
  • web comme completement nouvelle facon de creer puis d'utiliser la RV (rapid prototyping, link traversal)
  • travail de definitions et d'implementation de specifications (WebXR) comme dynamique complexe mais fondamentale (role du W3C)

Pour conclure le web n'est pas juste un Neme canal de distribution de nouveaux contenus de RA ou RV, il est deja le moyen le plus efficace de creer, tester, partager ce contenu. Il permet de plus des moyens d'interaction distribue et sans permission d'innover (permissionless innovation) qui sont si necessaire lors de l'emergence d'un nouveau medium. Finallement le web permet la connection de contenu et donc de createur de consommateur de contenu qu'aucune autre solution de distribution de RX ne permet. Le web sera donc le medium par lequel l'interconnection des contenus de RX laissera emerger un future que nous pouvons encore aujourd'hui difficilement imaginer.