The year is 2040. The web has evolved.

14 billion people have registered one or more pseudonyms, unique cryptographic usernames which serve as keys, allowing them to access and exchange their personal data. It allows them to be known by some, invisible to others. Rings of overlapping permission groups provide users with a granular mechanism for sharing content.

Everything is a file. Except there is no file, until there is. Hashing algorithms calculate arbitrary indices and regions within a file in real-time, enabling us to share arbitrary subsets and supersets of content.

These files exist as part of the same, universal, distributed file-system. In essence, a single, infinitely granular file with a complete provenance trail. Each file is version controlled, its previous renditions preserved in time. Many of these files have been annotated by those who are authorized. Annotations are not just multimedia, they are hypermedia. Inference, clustering, distributed analytics, it's all done automatically at the protocol level, no longer owned by an individual. Distributed search can be done automatically by inferring relationships between semantically tagged content.

PAuth2 (Pseudonym Authentication) has been streamlined through the use of these crytpographic psuedonyms. PAuth1 was complete failure. Passwords are no longer stored on disk, instead they are calculated by both parties as needed. The adoption of PAuth2 is as widespread as https in 2020 -- a RFC has been ratified by the IETF encouraging all services be interoperable, irrespective of their OSI layer or platform. API keys are no longer required. Like ssh-add, your system maintains a secure PA2 session which all servers can validate. What happens when you walk away from your computer? You don't, it's 2040.

Alerts and notifications cause chain reactions across services. All you have to do is set your preferences in one service and other services will be able to read them. Paying rent: automated. Daily meals, delivered: automated.

Project Diaspora, the decentralized Facebook, has re-emerged, built on top of our IPFS-like web 2.0 and PAuth2. It's has a popular node, like irc's freenode, but some people host their own private nodes because they are hipsters and whistle-blowers.

Google is more powerful than ever. With the rise of interoperability, they've been able to consume more data than ever. They have made several unsuccessful attempts at absorbing the Wikimedia Foundation to have tighter control over information availability. Google's search algorithms are still proprietary and now cost money, however they are interoperable and integrated at all levels of our stack, from word processing to video to chat.

Like bitcoin, people get paid to use their machine to store parts of the web and compute the hashes of files. The hashing algorithm changes weekly to deter those form monopolizing computations with custom hardware. Previous file hashes are combined and re-hashed. But FPGAs have become quite proficient and neural networks quickly re-learn new configurations which optimize for the hash computation. Inadvertently, opposite the original design goal, power has ended up in the hands of the even fewer.

In 2015 there was a web browser with an omnibar, it was a VM that lived on top the operating system. Today, there is a ubiquitous operating system. It is the file system. It is the web in 2040.

This is half ironic, half predictive and half, wait I am out of halves. Enjoy.

Original post

Related post on auth

Interoperability

What does a fully interoperable world wide web look like? See Reminiscing About 15 Years of Interoperability Efforts by Herbert Van de Sompel of Los Alamos National Laboratory and co-author of w3c OpenAnnotation