On The Value Proposition of Decentralisation
What Can We Learn from The Evolution of The BitTorrent Ecosystem?
Thanks for a very thoughtful response Denis. I’ve discovered that Medium doesn’t really distinguish between articles and comments in terms of formatting so I’m writing this as a new blog post :)
How is ‘decentralisation’ valuable?
I don’t think that decentralisation by itself is valuable to people, outside of a group of technologists, geeks and enthusiasts. Most of the time a decentralised system will be more difficult to use, have worse UX and in general will be less appealing to the general user.
For most people decentralisation will be an ‘invisible mean to an end’. Being able to have a video chat on Skype for free is definitely valuable but people won’t think about it in terms of a ‘decentralised tcp/ip network’ that enables that.
Free Skype calling is what a decentralised internet enabled but it wasn’t promoted as a ‘decentralised telephone’.
Likewise, Twitter is not a ‘decentralised newspaper’ even though it can be used as that.
Decentralisation is a property of a system that undoubtedly has many benefits. But these benefits can be understood by a narrow group of people.
For the rest of us, user experiences and value propositions built on top of decentralised systems is what matters the most.
That’s why I also think that ‘dapps’ or ‘decentralised applications’ is a great developer-facing term but will be confusing to the wider audiences.
Let’s look at the BitTorrent ecosystem to further investigate this issue.
The BitTorrent Example
From The BitTorrent Client to PopcornTime
BitTorrent is a great example of a most popular decentralised file transfer protocol which is responsible for a large portion of the Internet traffic.
While the network itself is decentralised and cannot be taken down, the discovery mechanisms for Bittorrent are very centralised.
These ‘discovery mechanisms’ are the PirateBays and other search engines for torrents which are called the ‘torrent trackers’. Leaving aside the questions of legality, these are the equivalents of Google in the Bittorrent space.
And we have the same economics here as we have on the wider Internet. Users host all the torrent files and exchange them between each other but it’s the torrent trackers that make most of the money.
Even though the low-level protocol is decentralised, aggregation & discovery is the most profitable endeavour which is always centralised.
Your attention is limited so you have to use a search engine, a catalogue or an app-store to find what you want.
The same will happen with blockchain apps whether that’s Ethereum or any other blockchain protocol. Discovery will be centralised and there will be someone making most of the money.
Interesting evolution of the BitTorrent ecosystem was the Popcorn Time application which abstracted out all the protocol-level features and presented users with a Netflix style, beautifully designed catalogue of movies. Available on-demand via streaming.
It looked like this:
Compared to that (the classic BitTorrent Client UX):
Again, leaving aside the legal matters, PopcornTime offered users a better experience of watching and discovering movies than the centralised Netflix.
There was no promotion of BitTorrent, decentralisation, torrents, magnet links or any other protocol level features. Instead, they:
- had an entire catalogue of movies and TV shows (all competitors are subject to licensing deals excluding certain titles from their offers)
- offerred a no-login experience —just open an app and start watching
- were completely free
- abstracted out downloads so that it felt like any other streaming service, just press play and watch
- made the UX was so good that you didn’t have to know that BitTorrent operates behind the scenes, you could just enjoy the movies
The decentralisation of BitTorrent was the reason all these features were possible. But users didn’t come for decentralisation, they came because the value proposition was better than the alternatives. (Sidenote: PopcornTime is now discontinued and developers are working on the ButterProject which is the PopcornTime technology without the controversial catalogue)
What Amazing Experiences & Value Propositions the Blockchain Apps Can Offer Its Users?
I think that’s the key question we all should be asking ourselves. It’s not really about decentralising banks, Google or Facebook. They work just fine in the contexts they evolved in. And decentralised banks, Google or Facebook won’t resemble their ‘centralised’ equivalents. In the same way, Skype is not a ‘decentralised telephone’ or Twitter —’a decentralised newspaper’.
Things like censorship resistance, immutability, resilience are very important from the systemic point of view. But they are not the benefits the end user will value.
To me, one of the killer user-facing features of the public blockchains is the experience of ‘ownership’ in relation to the digital objects. Combined with the ease of transferring this ownership it enables us to create amazing value propositions impossible before.
And because we discuss discovery in decentralised systems, I believe that ownership of one’s own social graph or link graph is key to solving that.
Blockchains, by their design, are the perfect data-structure to hold the public, immutable graph of links between any types of nodes (whether these are user profiles, apps, websites or other).
Using this hypothetical social graph, nodes can exchange both abundant ‘likes’ and scarce ‘value tokens’ (BTC, Ether or any other crypto-money).
Because of the public nature of this social graph, there’s no barrier to entry to creating the next Google or Facebook.
Users can easily switch to a different ‘discovery interface’ but because they ‘own’/control the private keys to the underlying nodes in the social graph and receive value directly on the blockchain, the switching costs to new interfaces are close to zero.
So what does it mean to the regular users?
It means that they can send and receive value directly in the form of links or likes. Basically, ‘making money while browsing Facebook’.
Or attaching small amounts of value to links/likes in order to get attention of other nodes/users. That’s what I mean when I say that the social graphs of social platforms and the financial graphs of blockchains will merge into the one network. Especially, when the scalability of blockchains such as Ethereum will reach the level of thousands of txs per second.
But even today the ‘Harvard scale’ equivalents of the decentalised Facebook can be experimented with.