Information Overflow

Computing

Each of us is walking around with a lot of data. Not only what is physically on our smartphones and laptops, but also what we’ve got backed up to the cloud. The cloud gives us automatic backups, and places to store content by the terabyte, but the current model by which we interact with the cloud is far from optimal.

As it stands currently you download a portal into your online storage, and view the content through that application. Dropbox gives you general file editing, and Google goes even further allowing you to do everything you could do in the browser in an application on your phone.

This model has a few key downsides:

  • You have information stored in multiple services, making it harder to search though and find.
  • Your interactions with your data is limited to what the service provides.

What is information overflow?

The cloud should serve as the base for all your computing needs and everything should always be synced and updated to the cloud. If you need to access a specific document or photograph it gets downloaded onto your hard drive when you open it. If there is not enough room you just remove a lesser used file, after all, its backed up in the cloud. No matter what you always have all your documents, pictures, books and files. You can make better use of your mobile devices as a result. Syncing preferences, accounts, passwords and everything else would be incredibly simple.

In this way, information (photos, documents, etc) overflows into the next largest container, with the cloud representing ground truth.

We already have a weak form of this, in which we have specific services that back up items to the cloud, or sync passwords among your devices. However this is not the same a syncing everything, automatically, always. That level of consistency across multiple applications and accounts would make programming easier for app developers, and make the user experience when switching devices simple. In fact developers wouldn’t have to worry about authentication, their own servers or syncing content. This would be a huge boon for developers and allow them to focus on what makes their application great. Solving this problem would raise the quality of applications, and make it easier for users to keep files in sync. Dropbox has made some really promising steps in this direction with the announcement of Project Infinite, but it is not at this scale.

Phones as an entry point

If the cloud were fully embraced there would be no need for anybody to own a computer besides their smartphone. This is starting to happen with Microsoft Continuum feature, in which a phone is used with a larger screen, keyboard and mouse as a full computer. You could push this further and stream down content needed, and outsource expensive computations to the cloud.

Phones are also getting powerful enough for most common use cases, allowing people to work entirely off their phones. What matters is the UI code and having an extra screen to use.

Logical Consistency between devices

If devices were fully consistent with each other a whole host of issues would disappear. People wouldn't wonder why files aren't syncing, or where something went. Version histories would be easier to keep without worrying about disk space. Users could have everything at their fingertips without having to consider where it's located or how to get it. This is incredibly important for user productivity and peace of mind.

What is stopping this from happening tomorrow?

Network speeds aren't fast enough in a significant number of localities to support such reliance on the cloud. Around 80-100 Mbit/s seems like a good threshold for the this to become feasible across rural environments. Certainly with Gigabyte networks speeds this would play out quite well.

No large player has emerged with standardized interfaces to various cloud providers that would allow the cloud to be treated as a utility. In short we treat the cloud as something that programmers deal with, when we could probably abstract things further for the end user. Storage is Storage and Compute is Compute. It doesn't really matter where it comes from, just how much it costs.

Update 11-10-16: Looks like CloudRail just announced a standard interface to a number of popular cloud providers.