Bitcoin exemplifies ideal online community

Bitcoin—a popular cryptocurrency—has something of a bad reputation. Anonymous and digital, it is a popular choice for illicit exchanges, particularly those operating out of the dark web, such as the infamous Silk Road. But Bitcoin is not inherently unsavory, just anonymous, and that anonymity is a major draw. Regardless of your personal take on Bitcoin and cryptocurrencies, the design of Bitcoin suggests a methodology to create communities rather than services—a design pattern that perhaps ought to be considered by other corners of the internet.

More than most services, Bitcoin relies on trust. The users must trust the algorithm underlying Bitcoin, blockchain, to have any faith in the value of the currency, but there also needs to be an aspect of trust within the community. A paper by Sas and Khairuddin surveyed the Bitcoin community—and it is indeed a community—about why they used Bitcoin and how they related to it. In all their interviews the topic of trust kept reemerging. People expressed a certain trust in blockchain, either a faith in this particular use of it or just the algorithm in general. People expressed trust in the creator of Bitcoin and the vendors that use it. But interestingly, people most frequently reported putting trust in particular users of Bitcoin, individuals seen as knowledgeable and influential by the community itself.

The idea of placing trust in an algorithm is an important one, and one designers should internalize, but it can be somewhat abstract. When we say we trust something, it usually means that we expect a certain range of behavior. This is often rather general, e.g. I trust that my family and friends are not trying to hurt me. But matters are complicated by the mutability of algorithmic services despite their typically narrow set of objectives; that is to say, algorithms are typically useful for specific tasks where they constantly iterate on and tweak parameters to optimize performance. The algorithms we interact with regularly—Google, Facebook, email filters—are not static. Their workings change daily, sometimes slightly—like altering a filter on your Facebook feed—and sometimes dramatically—such as the roll out of new features like facial recognition.

Since we cannot place our trust in the ever-changing codebase of an algorithm, we must instead trust its behavior. We then believe that it will continue to function in the way that it has, or that it will improve features without bringing harm to the user experience. This type of trust is very akin to faith; we users can do very little to affect change on the algorithms we use. The only situations where we have bargaining power is in our collective outrage, which seldom does more than erase a few poorly thought out design choices and cause a series of lukewarm apologies.

Bitcoin does something different, though: In its crowdsourced nature it has effectively created a community of users with a sense of agency. From a design perspective, this is both completely obvious and yet strikingly elegant, but it does require something that most algorithms are unwilling to have: transparency.

Bitcoin can create this sense of trust because, while the individuals using Bitcoin are anonymous, blockchain, the ledger of all Bitcoin transactions, is completely visible. The visibility of this ledger is how Bitcoin functions, as users donate their computational resources to assess the validity of transactions and ensure that money is not being stolen or duplicated through poor record-keeping. In exchange for donating their resources, the user can be awarded some number of Bitcoins that are given out at decreasing rates. This process is known as mining, and the miners who are the most dedicated are often viewed as pillars of the community.

This works because blockchain is transparent by nature, but in many cases large algorithms are almost inexplicable. No single person will be able to tell you how all of Google works. This is in part because of intellectual property rights, but mostly because no single person will understand it completely. Google operates at such a large scale and iterates so rapidly that the algorithm never stays static enough for an individual to grasp the workings of it.

Most of us do not feel any particular attachment to Google, though we all use their products regularly. Similarly, despite their size, there is not a community of users built around the use of Google services. Consider Facebook: It is built around creating a sense of community, but that community is built entirely out of existing social networks. As much as Facebook likes to portray itself as connecting people, it mostly connects people who were already connected.

Similarly, I doubt any of us feel particular loyalty toward Facebook; it just happens to be the most convenient.

This model cannot possibly work for Bitcoin, nor should it. There is far too much competition—DogeCoin, PotCoin, Etherium, to name a few. Burger King has even created WhopperCoin. It is not just a good thing that Bitcoin operates on trust, it is essential. Transactions depend on the trust between people, the belief that individuals will render services promised based on goods received. Trust is essential in all our commerce, as we all use fiat currencies that are not inherently valuable—though we simply do not have time to dive down the rabbit hole of what gives objects value in the first place.

But from a design perspective, this sense of community and trust should be more present. People certainly form communities online— places like Reddit are a good example—but that is because community-building and interpersonal interaction is specifically fostered by these services. What would the internet be like if everyone had a stake in the various services they used, if we found kindred spirits based on browser clients? It could be an absolute nightmare, but maybe, just maybe, it would bring a few of us out of seclusion. And if nothing else, it would be a thoroughly interesting ride.

Leave a Reply

Your email address will not be published. Required fields are marked *