Decentralized Storage: Why I’m skeptical

Bedeho Mender Blocked Unblock Follow Following Oct 28, 2015

Let me preface this by saying, no one would be more happy if decentralized storage projects were very successful, I’m just skeptical that they will.

There are a number of projects which are working, either entirely or in part, on decentralized storage: Storj, MaidSafe, Filecoin, Sia. This means that data is stored on an open group of computers which are incentivized to perform this service. There are many obvious technical hurdles to overcome in making such systems work, but this is not why I am skeptical.

The key value proposition of such systems over traditional cloud storage is that by using underutilized idle disk space in the swarm, the system will be able to provide cheaper storage. Other benefits are some times also invoked, like improved privacy through client side encryption, faster download speeds. Even if we grant these benefits, there is an obvious question:

Why hasn’t someone already done this before, but centralized?

The fact is, the benefits of distributed storage as stated above does not require the technical breakthrough brought about by Bitcoin/Blockchain technology. There is no reason one couldn’t run a central server which coordinated the activity of peer computers, and rewarded them with a centralized credit system which could be redeemed/bought for fiat money. In many ways it could be better. Micropayments inside a single system are indeed not a unique feature of Bitcoin/Blockchain technology, PokerStars has been doing it for years.

So if such systems have been possible all along, but no one bothered to build it, then that is a very bad sign. New cloud storage startups are popping up all the time, and they are all opting to not go this route (distributed storage, centralized payments and coordination), that makes me skeptical.