Platform or Protocol? Chicken or Egg?

Very thought-provoking stuff thank you (finally some real content!)

I think the core issue as you expressed is one of network effects.

If this idea is to work how do you build to critical mass?

The wrinkle here is whereas other firms/organizations can typically focus on building either a platform or a protocol, you are (rightly so I believe) espousing that you actually need both more or less simultaneously.

The popcorn time analogy is a very good one. In that case, the protocol is torrent and the platform is the discovery layer that popcorn time layered on top of it. Arguably (putting the controversies on IP aside) the “total package” was not ready to go mainstream until both were mature.

The question for you now is how do you tackle the chicken and egg problem when each piece, both the protocol and the platform, is equally perplexing? Having the protocol alone would not be intuitive enough for most users without a clean UX on top of it. Building the discovery platform is a non starter without the underlying “decentralized” digital ledger.

So what to do?

One thought I have (just brainstorming) is to take a vertical slice rather than a horizontal one.

In the other piece you wrote on your blog, I think messaging is a great place to start. Perhaps a combination of a personal payments (a la square cash) and messaging could do the trick?

Given how saturated messaging is, this could be a difficult road to climb without introducing something unique on the messaging side of things (like sending cake?!).

Another thought is to coopt an existing discovery platform by layering a ledger protocol beneath a twitter or a facebook. However, in this case you lose one of the core tenets of your proposal which is to decentralize “ownership” of the graph. The tradeoff is this may be a way to bootstrap and jumpstart the protocol side of things and worry about the discovery UX later.

Finally, here’s a fun one to think about, what if there was a Medium competitor that leveraged this protocol for both discovery and “transactions” (both monetary and attention based in your parlance)?

There are a ton of complaints currently that the discovery mechanism in Medium is broken. My argument is correcting for this is not a quick fix, it is unfortunately endemic to the way Medium is currently structured, i.e. by attempting to mirror facebook and twitter’s like and follow paradigms.

By attempting to build their own proprietary graph on these principles, they are running into a lot of growing pains, not least of which is a growing dissonance between what their algo thinks is quality writing vs what actually resonates with people. Secondly, they are still in effect “owning” the content without “owning” the content, i.e. they own the graph which these days (to your point on content vs attention) is paramount to owning the IP or at least the ability to find it.

What if, instead, there was a platform where content creation was as clean and intuitive as it is on Medium, however, the underlying system for discovery and propagation was based on your generalized blockchain concept?

In this hypothetical scenario, content writers have a much better chance of monetizing their content (even if in small amounts, pennies on the dollar if you will). They would benefit three fold:

• They own the graph as well as the content
• They can monetize directly at fractions of a penny (reduces barrier for readers to contribute)
• They have a better chance of being discovered which doesn’t rely on “brute force” of # of followers

In this context, rather than rely on a system of “likes” or “recommends” to determine quality, you could have a must more robust analysis of the historical relationship between the contributor and the patron. You would also, to your point, have an idea as to the recursive importance of each party (they can effectively pass on their attention index to others).

This would provide a much more powerful and flexible mechanism of discovery. Furthermore, imagine a scenario where you can open-source the discovery mechanism?

Part of the problem with Medium right now is they, by definition, need to tweak the black box until they get it right. They have to effectively build the perfect discovery engine simultaneously for all users and each user.

This is near impossible. No wonder they seem to be swimming upstream with their users lately.

With the blockchain mechanism whereby you separate the ledger from the discovery, you could build the “original flavor” of discovery and then open it up to all comers. Letting the community determine organically which modes of discovery and weightings best serve them.

The final consideration here is to make sure that new users have as good a chance to be discovered as others. Again, there’s a couple ways to approach this:

• Open source discovery allows others to create specific algos to “find up and comers”
• Weight “velocity” of transactions (100 transactions in 24 hours) with a similar magnitude as “reputation” (1 transaction from a party that has 1 million historical transactions).
• Consider weighing contribution relative to reputation, (a transaction of value “1.0” from a patron with a historical reputation of “0.1” would be weighted much more than a transaction of same value from a patron with a historical reputation “10000.0”). i.e. if a patron punches above his or her weight class to show you appreciation that is worth more than a heavyweight throwing a couple pennies your way.

Anyhow, here are some thoughts, happy to brainstorm further with you, I find this idea intriguing and frankly (having quit my prior job 4 months ago) I am eager to find some ideas that are worth building out further.