We Have Some Blocksplaining To Do

Randy Meech - July 16, 2018

The open mapping community is skeptical of blockchain. While skepticism is healthy and there's plenty about this field that calls for it, too much skepticism is unfortunate. At its most pure, blockchain technology and accompanying crypto tokens provide some of the best incentives and governance potential for completely open software and data. Despite all the hype and nonsense at the moment, I believe that in a few years the center of gravity for open source will move to support blockchain projects.

Mapping, of course, is hard and requires a lot of expertise. Since blockchain projects are inherently open, they need strong, supportive contributors if they're going to work. Since there's a disconnect, we need to do a good job explaining what we're doing and why we're doing it.

StreetCred recently started to talk more specifically about our work, starting with a Techcrunch article, "StreetCred is building a blockchain-based marketplace for location data." We got lots of feedback! One of my favorite responses was from Tom Lee of Mapbox, who has a great understanding of open data projects and their governance:

I am pretty skeptical of blockchain applications *except* as an institutional commitment mechanism, where it might have real potential. I think this is how @randyme's new work on crowdsourcing POI geodata is approaching the tech: https://t.co/15Pb3vTqEP

— Tom Lee (@tjl) June 21, 2018

Of course, this was followed immediately by Nelson Minar, who taught me all about APIs when we were at Google about fifteen years ago. In case you can't tell, his question is not really a request for information :)

When is the ICO?

— Nelson Minar (@nelson) June 21, 2018

Adding insult to injury, Burrito Justice (formerly of Mapzen) made fun of me when I commented on his dinner-grilling setup:

cooking it via blockchain to everything should be ok

— Burrito Justice (@burritojustice) June 27, 2018

Ian Dees, one of the most prolific developers working on open addressing and open POI data, actually liked our omission of the word "blockchain" in an early meditation on a blockchain protocol:

"blockchain" is a loaded and distracting term nowadays, so it's refreshing to see @randyme imagine a protocol to share POI without mentioning the word blockchain: https://t.co/pkxsHJPhtb

— Ian Dees (@iandees) April 12, 2018

Most recently, Chris Holmes commented in a similar vein on the same topic:

I missed this article the first time, but it fully aligns with where I'm at. It's refreshing to get all the way through an article about decentralization that doesn't talk about cryptocurrency or coins at all... (outside of one small link out).

— Chris Holmes (@opencholmes) July 16, 2018

I think it's safe to say that we have some work to do to describe more about what we're doing at StreetCred in the clearest possible terms: blockchain, tokens and all.

With apologies to my Latin professors...

StreetCred Mappa de Locis Blockchain est

There are intractable mapping problems for which blockchain offers bright promise. StreetCred is focused on points of interest: the restaurants, doctor's offices, and other places you care to visit. Even with the highest-quality datasets, we've all used an app to find a place to learn too late that it closed. This is due to an incentive problem: what's our incentive to mark something as closed in an app that's just disappointed us with incorrect data? Points of interest are more dynamic than other types of map data; they show data problems more quickly than roads or bodies of water, which change infrequently. They're the hardest problem in open mapping, and an ideal candidate for a new incentive structure.

Building Incentives for a Billion POIs

Yes: we believe there are a billion POIs in the world, or ten times what both Google and Foursquare have in their datasets. Why such a big discrepancy? Previous incentives have included direct payment and in-app game dynamics to collect data. Neither of these has scaled globally, creating large, underserved geographies. Even in places where it works best, it could be better: I see problems in Brooklyn, NY all the time. We want to represent the world the way local communities see themselves, and are working on incentives to make this possible.

First, StreetCred is building tools to analyze and predict the quality and coverage of POI data across the world. We'll be publishing more about this soon, and it's critical as we develop incentive systems that will get us to our billion POI goal. I've personally needed tools like this across many jobs, and I can't wait to share them with you. When we're done, you should have a better understanding of what POI coverage should look like in the areas that you care about. If you're in the mapping industry, these tools will help you understand the data your POI vendors are selling you, and how StreetCred data can help.

It's critical that anyone can contribute data easily. One of our top goals is to reduce complexity for would-be data contributors, especially entry level ones. Many open data communities suffer because new contributors find it too hard to participate. For a blockchain project, this will be especially challenging because of the inherent complexity of managing wallets, security, devops requirements, etc. Limiting data contribution only to crypto-enthusiasts is a sure way to create a badly misaligned, non-representative dataset, and we will avoid this at all costs. Instead, we'll offer an easy-to-use reference application built on the protocol, and invite others to do the same.

Given how open we want to be toward data contributors, we must offer strong, in-person validation. This is a great use of blockchain technology and smart contracts. Data contributors should be rewarded when they create good data, and not rewarded otherwise. StreetCred validators operate in a way similar to projects like Wikipedia or OpenStreetMap, where experts care for their areas of interest and expertise. As the value of the dataset grows, we hope that our validators will share in that value by participating in the protocol.

Here Be Tokens

Tokens are critical to the StreetCred protocol, and the overall token economy we're creating will reflect the underlying value of the dataset as it grows over time. Most importantly, early participants should be rewarded for creating valuable data. We believe that global data collection will scale rapidly if the incentives are right and data quality is high.

Collecting all the POI data in the world has never been done before, mainly because it hasn't made economic sense. Why would a company pay people for POI data in a region where it's not economically viable? Utility tokens flip this on its head: of course there's value in data about a place, even if it's not as marketable as data about Manhattan. We know there are organizations that care about data all over the world. If there were a way they could incentivize that collection, they would do so. Blockchain projects cut out intermediaries by definition. By removing the need for a company to seek rent on overhead, a protocol can pass incentives from those who want the data to those who can create it, opening up entirely new possibilities for data collection. I'm most excited by the prospect of bringing economic incentives to places that are now literally not on the map.

Because of our background in commercial mapping, we know most of the companies who buy this data currently. It's an understatement to say that none of them are happy with the way things are with the current data providers. To (anonymously) quote an industry veteran: "Your favorite POI vendor is the one you didn't go with last year when you signed with the other one." We are working hard to make current data consumers an early and important part of the StreetCred protocol, knowing that their demand for data will fuel its development. Transitioning companies into token consumers will be challenging, but it's some of the most important work at hand if we're going to succeed.

Ex Crypto Claritas

It's safe to say that the blockchain space has a communications problem. Part of the problem is a funding-driven pressure to speak in byzantine, technical language in advance of working products with engaged communities. But blockchain projects also need to cultivate developer networks, so complex communications are counterproductive. We need to do a better job of showing utility tokens as a promising tool for anyone who cares about this kind of data. There are so many who care about open data who are turned off by the default way of speaking about blockchain.

StreetCred comes out of a decade of work on open mapping problems. Our overarching goal is to collect the world's billion POIs in all their dazzling diversity. We care a lot about this. To succeed, we need to be clear and open. This is what you'll get from us through our words and our work.