#25 What prevents a truly decentralized digital world
This week was about incidents related to platforms blocking access or abusing their distribution powers. GitHub (rather Microsoft) took down Youtube-dl, the most popular youtube videos downloading software on a complaint from RIAA. Facebook sent a cease-and-desist letter to NYU researchers trying to gather data around its political ad targeting systems. Zoom apparently deleted a scheduled Zoom meeting that discussed censorship on Zoom. Apple in Hong Kong refused to engrave ‘Liberate HKERs’ and announced their decision to do so over an informal sneaky phone call instead of an email. Twitter locked down a newspaper’s Twitter account, only to reverse its decision later.
Complexity is the price of decentralization
The biggest impediment to decentralization lies in the design of things making it so much more difficult to use. Vanilla Bitcoin, by design, is not instantaneous like a card payment. I am supposed to know stuff like this but if someone asks me if there is a way to publish a piece of document in less than a minute on the web so that no one can take it down, I might not be able to reply with an acceptable answer. It is a miracle that torrents have survived as a technology of file sharing in a world where the movies and records industry goes all guns blazing after everyone who pisses them off, even indirectly.
There is no shortage of decentralized alternatives for users. There is a decentralized video platform (Peertube), a decentralized marketplace to buy and sell goods online using cryptocurrencies (OpenBazaar), a decentralized social network (Scuttlebutt), a Dropbox alternative incentivized using crypto tech (Sia), tools to roll out your own decentralized communication platform (Matrix), even an alternative to the current World Wide Web (IPFS). I have not used any of them ever. Most of these complex issues are difficult to overcome unless decentralization becomes part of the web specifications itself.
Filecoin, a storage application built on top of IPFS adds an incentive model to allow people with unused storage space to rent out their space for files of people who are willing to pay for such storage services. It uses a cryptocurrency based model which though technically sound, has been subject to politics, miners revolt, and strikes affecting stability in the past.
Complexity also arises in maintenance and administration. Large tech corporations love to have control over the various parameters that govern the quality of service. It is much easier to do so with machines and compute capacity that is entirely owned and controlled by them rather than decentralized (though they love to do compute distributed over various machines that they control for reliability, there is a difference). There is an added legal angle wherein a government might just stop operations of a tech service for not abiding by any takedown requests. The argument of inability due to decentralization does not hold even in many functional ‘democracies’.
Decentralization is technically complex too. There are multiple additional things to take care of. Availability for example in file storage systems becomes a concern. Controlling latency is another. So much so, that systems like peerTube degrade poorly when peers with slower connections are added to the network. Proper incentive design and preventing abuse and price manipulation leading to a denial of service is yet another issue.
Having said all this, I still believe that decentralization is the way forward in the coming decades. It just needs one big push to become mainstream, something like what Bitcoin did for crypto.