Inverting the Internet

Smartphones and the cloud together have unleashed a abundance of devices and applications that have become integral to modern life worldwide. These devices and their applications in turn depend upon a massive amount of data stored in the cloud by application. This silo approach to information reflects the state of computing ten years ago and contractual restrictions on Apple (which by serendipity I wrote).

This creates two problems which are looming ever larger as time goes on. First, any application becomes a target for hackers, and whether from hackers or error data can be wrongly exposed. The problem is so large the European Union is promulgating the Global Data Protection Regulation effective May 25, 2018. Second, new classes of applications, many built upon an artificial intelligence framework, require access to all of a person’s information to be fully effective. There is just no way to do this easily.

We need to invert the internet. Index by person rather than by device and/or application. Put all the information within an encrypted container for one person. The owner of that container can then decide what information to share with whom on what terms. At the bottom layer are countless instances, one for each person, within which run the applications and services desired.

Any breach now will be of a single person. All data is encrypted. GDPR safe harbor provisions are met. If a government has a warrant to access a person’s data, it can gain access to it, but otherwise no one else can including the service provider. Intelligent agents whether Bixby, Google Assistant, Alexa, Cortana, Siri or one of many others can be granted access by a user to whatever information they wish, but that information is processed anonymously within the container.

Every application the user runs today on the smartphone still runs, but the application no longer knows or saves the information which it is processing, just its results. Streaming services for music and video run better in this model than today because network paths and caching can be better optimized. Even picture taking is little changed as both Apple and Google upload pictures in real time as they are taken. Devices become networking and sensor hubs, but some local applications likely will remain.

Thin client devices are possible and they may be legally mandated in some circumstances. By their widespread adoption of VDI, corporations worldwide have shown their need to protect their data on corporate servers. Already, in the U.K., financial institutions have been directed to move to mobile devices where the institution can monitor and archive all data. This architecture meets that requirement.