Block by Block: Data Storage

Block by Block is a series where we dive into different industries and examine the entry-points for decentralization.

Data is becoming the world’s most valuable resource. In the Information Age, companies that control the ebb and flow of data are the power brokers. As humans continue to generate data — 2.5 quintillion bytes per day — deciding how it’s stored becomes increasingly important.

As the shift from on-premise data storage to cloud storage continues to accelerate, tech giants have begun to stake a claim in this $30b market. The ability to access your data and your files anywhere you want is massively beneficial in increasing productivity. However, this benefit is not without drawbacks. In a world where data is stored and managed by a handful of companies, the risk of centralization becomes apparent. With hacks and censorships becoming alarmingly common, new decentralized alternatives are popping up. Research from The Block, show that at least 30 projects are attempting to build decentralized data storage platforms.

Where are the points of entry for decentralized data storage platforms?

• Central points of failure: Centralized data storage platforms are, by their very nature, also central points of failure. As such, these platforms are prone to hacks and server crashes. In 2014, hackers were able to access private and intimate photos and videos of celebrities by illegally accessing their Apple iCloud accounts. In 2017, a typo in the Amazon S3 system took down the entire system for nearly five hours, impacting multiple platforms such as GitHub, Slack, Twitch, and Expedia. Using centralized data storage forces users to trust that these platforms are reliable and that their data is secure.
• Censorship: Centralized data storage platforms can cut off user access to their services in acts of censorship. In 2016, Chinese data storage providers were forced by the government to cease storing certain images and videos to create a "clean" internet environment. In 2018, Amazon threatened to suspend the AWS account of anonymous messaging service Signal over what Amazon believed were acts of censorship circumvention. The threat of censorship is a constant risk to businesses and users of centralized data storage platforms.
• Oligopolistic: In the first quarter of 2018, just four companies (Amazon, Microsoft, IBM, and Google) controlled 60% of the cloud infrastructure market — Amazon alone, has a 33% market share. If a handful of companies continue to maintain their dominance, oligopolistic features — barriers to entry for startups, price control, etc. — will become more apparent.

With the problems facing centralized data storage platforms, how do decentralized data storage platforms solve them?

• Central Points of Failure → Distributed Servers: Instead of a handful of server farms storing data, decentralized data storage platforms run on a large distributed network of servers. If one of these servers is attacked or shut down, data can still be retrieved and access from other servers on the network.
• Censorship → Uncensorable: On decentralized data storage platforms no single entity can dictate the removal of content or prevent access to data storage services. Furthermore, because no single entity controls data accessibility, decentralized platforms cannot be pressured by governments or corporations to censor specific individuals or businesses.
• Oligopolistic → Competitive Market: A large distributed network of servers creates a competitive market. On decentralized data storage platforms, server operators compete on price to receive to right to store your data. Competition creates downward pressure on the price of data storage. As an example, as of September 2018, Sia has 590 storage providers on its platform. Data storage on Sia cost ~$2/TB/month compared to Amazon S3 which charges ~$30/TB/month.

What are the barriers to entry?

• Reliability and Scalability: Because they are in their early stages of development, decentralized data storage platforms have yet to prove that they are reliable at scale. Moreover, some tests suggest that the cost estimates to use decentralized data storage platforms are unreliable and often inaccurate. Unreliable accounting and unproven scalability could deter the adoption of decentralized data storage platforms, especially for large enterprises.
• Legality: There is an ongoing concern about the risk of storing data that might be considered illegal. Because they are decentralized, these platforms should, in theory, be unable to censor specific content — even if the content runs afoul of the law. This raises the question whether data storage providers can be held accountable for what users are storing. If so, the legal risk could prevent storage providers from adopting the platforms. Currently, developers are introducing ways to allow storage providers to “graylist” certain content, preventing their servers from hosting it. A graylist, however, could ultimately lead to less popular content -- even if it’s perfectly legal -- from being shut out of the platform. That effectively ends up censoring said content.

Decentralized data storage platforms are crucial for the decentralized economy — adding the infrastructural backbone for decentralized applications. In an increasingly digital world, the power to securely store and access your data will soon be an essential part of modern living.