Economic Data Siloes

Data Siloes

In the centralized web, consumers’ data has become siloed amongst a handful of powerful players centered in Silicon Valley. In other words, FaceGoo vacuum up staggering amounts of data pertaining to online behaviour - and most aspects of most First-World people’s real lives are reflected in their online behaviour.

All this sensitive information sits on their computers, and you can’t erase, or see it. If you would like to get a copy to move elsewhere, you can’t. For instance, many small businesses or minor celebrities or whoever write a lot of posts on Facebook.

You can’t get a copy to put elsewhere; you have to find and copy it. Your data sits in a bunker, in a silo: it is siloed. And if Facebook’s servers got hit by a meteor, all your memories would be lost.

The same worry applies to much other important information: scientific data, economic data, records of works of art, government documents. It all resides on private servers, hard to share, hard to access, vulnerable to accidents.

Decentralized Web and Future-Proofing

In the decentralized web, we split data up amongst many participants. Because the data is split up, no one computer can take it down. Because it’s practically impossible to control that many computers, no one authority can control the information. (Under the EU’s right-to-be-forgotten policy, Google has been forced to censor search results) This decentralized protocol prevents technical data siloing.

Information, like books, shared on a decentralized network will persist long after a corporation, like Facebook, goes out of business. It will persist as long as participants on the network keep the information alive by sharing it. This is a hopeful thought, because it implies the information will die out only when no one wants it anymore.

I love the Long Now Foundation. They try to think, in detail, about civilization over the course of ten thousand years, going so far as to write dates that way: it is 02017 as I write this. They are building a clock which should last ten thousand years, and they intend to be around to steward the clock.

Is a single foundation the best way to future-proof a clock, or to future-proof their library, The Manual for Civilization? I wonder if an economic protocol incentivizing the safeguarding of these assets might help.

For instance, Swarm incentivizes data storage and retrieval. It has its own token. Wolk and several other new technologies are implementing database APIs on top of Swarm, incentivized by their own tokens.

Economic Data Siloes

Suppose we had numerous networks like Swarm, each of which stored and retrieved data, incentivizing and returning data through utility tokens. That is, those who store data get paid in this monopoly money to do so, and without the token one cannot send data.

What if you want to transfer information from one network to another, though?

There could be a technical barrier, in that you or someone has to be a member of both networks, able to engage in both economic and technical protocols.

I am thinking of Polkadot and its scheme of a hub-chain and parachains. It aims to be a conduit between chains, able to take transactions and tasks and transfer them to other chains. For instance Ethereum computations could be put on another smart contract network to be run, and the result transferred back to Ethereum.

With a proliferation of data storage protocols, we may need to devise a similar hub protocol for documents. Concretely, who will translate between Swarm and IPFS? In the end, though, such technical details probably won’t matter too much.

However, transferring information metered by tokens could hit an economic barrier.

What if your information is stored on a dying network, and you want to get it out? As long as the retrieval costs are low this is no problem, but what if they were high? Your information would be expensive to retrieve; it would be siloed again, locked in by an economic rather than technical barrier.

This probably will not happen, because if the network is dying then its token value will drop, and getting out the information should be cheap. Take the opposite approach: suppose you stored a great deal of valuable information on a network (which, again, charges to read the information), and the token value went to the moon.

I will be keeping a close eye on the possibility of economic data siloes, a potential emergent property from many token-metered systems. Swarm, IPFS, Ocean Protocol, and the other startups currently working on information sharing have the potential to unlock a great deal of value by breaking open old data siloes. It would be a pity to build new ones inadvertently.