Something is deeply wrong with how social networks on the internet are working. Whether it’s the steady stream of revelations about how our personal information is harvested and sold for profit, how hate speech is allowed to flourish, or how the response to hate speech has been to deprive people of their right to speak, there’s hardly anyone who feels that our social networks are working in their best interest.
Technological Solutions?
Today I want to talk about the technological solutions we at Steemit have developed, and continue to develop in an attempt to solve these problems and move us toward a more open, decentralized, and free future. A future in which people are motivated to engage in constructive conversations as opposed to sacrificing their privacy for the privilege of engaging in angry and divisive screaming matches.
These views are my own and do not reflect the views or beliefs of Steemit Inc. This is also not investment advice.
Hate Speech
Given recent events, it seems reasonable to focus on hate speech and how the issue is currently being handled by those organizations people have been using for years to act as the custodians of their public and private information. It should go without saying that nobody wants their social networks to be a fertile breeding ground for evil individuals or organizations. The response of the major social media platforms has been effectively to create internal committees that operate in a top-down manner to ban those accounts they believe to be producing Hate Speech. This is what we call a “centralized” approach.
The Problem of Centralization
Believe it or not, one of the keys to Facebook and Twitter’s success is their centralization. Storing the social information of billions of people requires an incredible amount of infrastructure which costs an unfathomable amount of money. Maintaining total control and ownership of that infrastructure was key both to scaling their operations as well as generating revenue through the sale of the information stored on that infrastructure. For now let’s put aside the moral hazard that results from an entirely ad-based revenue model and maintain focus on this issue of centralization.
The Benefits of Centralization
This centralization is what enabled these organizations to scale so rapidly while also constantly improving their product-market fit through A/B testing and frequent iteration which, in turn, helped make their platforms literally addictive. As more and more people wanted to use their services for an ever expanding percentage of their day, this centralization served dual-duty by helping them to scale their infrastructure as rapidly as their growing usage demanded. It also allows them to respond to threats fast. For example, when there is dangerous hate speech on their platform, they can quickly remove the speech, and the person, whenever they want.
The Fatal Flaws
There are, however, three serious problems with this approach: 1. The only people worth removing from your platform are those with significant followings, 2. that extreme rhetoric is a good way to grow a following on platforms that focus on eyeballs as the key metric as opposed to value, and 3. that removing anyone with extreme views and significant followings only turns them into a martyr and empowers them to promote a narrative of persecution and unfairness. This is an attack vector that centralized organizations are not capable of solving, which is why the situation continues to get worse and worse, with no end in sight.
The fundamental problem with centralized systems is that they are extremely difficult to make provably fair. But, as we’ll see, the only way to do that is to become more open and decentralized, which would conflict with everything that makes these internet giants successful.
The “Hands-Off” Approach
One solution to this problem of hate speech is to build a platform that promises a “hands-off” approach to censorship. But a promise to not regulate speech comes with its own set of problems as exemplified by Gab.com, which as of the time of this writing has been made inaccessible, presumably due to the fact that it was used by a recent shooter to publish certainly angry and hateful speech.
Gab.com
Gab’s solution is yet another centralized platform, but with the addition of what is effectively a promise not to censor speech. That might sound good to some on the surface, however, when you think it through it’s equally untenable. The value of social networks comes from network effects. Facebook doesn’t offer any especially revolutionary features, aside from the fact that practically everyone on Earth is already using it. That’s precisely why it’s the most valuable social network on the planet. That means that if a person is going to join a social network, all things being equal, they will derive the most value from joining Facebook.
Attracting Deplorables
If you offer a “Facebook or twitter alternative” premised on the fact that you will not censor anyone who uses your platform, then your network will only be able to provide more value than Facebook to people who are incapable of using Facebook, like those who have been banned. While I’m sure there are some good people and outlets that were collateral damage in what’s been called “Facebook’s Purge,” the group as a whole is probably not one that most people are eager to join. That’s why the majority of these platforms inevitably become a safe-haven for hateful rhetoric. But if the solution is not centralization, or simply promising not to censor any speech, what is it?
Steem is the Solution
The solution Steemit pioneered with the release of the Steem blockchain in 2016 is one that leverages new decentralized technologies (like blockchain) to create a provably fair and transparent system that empowers the crowd of users themselves to self-regulate what content is hidden from view, and what content is financially rewarded. On the one hand, everyone is free to store their text-based speech on the Steem blockchain which makes it effectively impossible to censor.
The key distinction lies between information that is stored on Steem, and information that is displayed on Steem interfaces like steemit.com which has become the most used blockchain-powered interface on the planet. While anyone is free to store their speech on Steem; any Steem developer is free to display‒or not display‒that information on their site. This constitutes yet another layer of decentralization.
Multiple User Interfaces = Layers of Decentralization
While someone’s speech might be censored on one interface, there are now many other interfaces available to that user like busy.org or steempeak.com, which will be regulated by their own community’s social norms. An added benefit of this decentralization is that if one platform is taken down, or simply breaks, users will still be able to access their content through other platforms. The fact that Gab.com has been inaccessible for days proves the dangers associated with centralized control over free-speech platforms. Had the people at Gab simply built their user interface on top of Steem, their innocent users would still be free to access their content and engage with their followers on other web interfaces like steemit.com, busy.org, and steempeak.com. At the same time, the users of those interfaces would have the ability to regulate any bad actors that might have been attracted to that platform by downvoting their content.
But what if none of these communities using these varied interfaces want specific speech to be displayed? Well then, it probably is hate speech! But even that speech will still be preserved on the blockchain, and anyone who believes that speech is being unreasonably censored on every existing interface is free to launch their own, and display that speech. Because Steem is the only blockchain protocol built from the ground up to power web applications, developers can leverage all the mature and state-of-the-art web development tools that have been created over the last 20 years to rapidly launch their own blockchain-powered interface in days as opposed to years.
Cryptoeconomic Regulation of Speech
Probably the most important, and underrated, aspect of Steem’s regulatory mechanism for hate speech is what we call “Proof-of-Brain.” STEEM tokens are the native currency of the blockchain and these tokens are distributed entirely based on the “upvotes” and “downvotes” of the Steem users themselves. Effectively there are two separate capabilities relating to free speech embedded into the cryptoeconomics of Steem. There is the ability to post whatever information you want to the blockchain and there is the ability to earn valuable tokens (STEEM) in exchange for sharing that information.
If you get enough upvotes from enough people, you earn STEEM tokens which have been valued as highly as $8 a token. If you get enough downvotes, you won’t earn anything and interfaces like steemit.com will hide your content from other users just as it does with spam. In fact, from the perspective of the blockchain, there is no meaningful distinction between hate speech and spam. Both are just content that has no value, or even negative value, to users.
Rewarding “Ordinary” People
While the ability to deprive people of rewards is certainly an effective regulatory mechanism for encouraging good behavior, what’s more important is that there is a strong economic incentive to create sufficiently inoffensive content. This is what other centralized “free speech” platforms are lacking; a system for providing value to ordinary people value beyond that which is offered by Facebook.
What many of the newest members of the free-speech-bandwagon fail to appreciate is that the right to Free Speech is premised on the fact that we are all created equal. Every voice matters just as every life matters. And we believe that everyone deserves the opportunity to earn value for sharing their thoughts. That includes Democrats, Republicans, Moderates, nice people, not just the people so angry no one else wants them. Ironically, welcoming everyone into an open attention economy, is the only way to protect free speech.
Hate Speech is Not More Valuable Than Peace Speech
To put it quite simply: everyone has to have an equal opportunity to earn rewards for their content based on its value as assessed by largest number of people. The mistake that most “free speech” platforms make is that they assume, consciously or not, that the speech of banned people is inherently more valuable than the speech of others. Otherwise, they wouldn’t design and market their product to appeal to that demographic.
On the other hand, what would happen if you created a platform that attracted a diverse user base, from all walks of life, each equipped with the ability to vote on whether content is valuable or not (like hate speech)? What kind of content do you think would be rewarded, and would be hidden? Whether you agree with this system or not, no one can deny that it is provably fair, and that is where its strength lies. Everyone has the ability to vote on content and everyone has the ability to increase their stake if they want more influence over who gets rewards and the size of those rewards.
Ad-Based Revenue
My goal was to give you a better understanding of how websites are currently struggling to deal with the issue of hate speech, and explain how we developed Steem to address the problems with their approaches. Another big problem with existing platforms is that they are entirely dependent on ad-based revenue which means that they have no choice but to turn their users, into their product. I hope to explore that issue, in a future video.
Thanks for reading.
If you're attending SteemFest 3 next week, I will be hosting two panels. On Day 1 I will be interviewing Steemit CEO, Ned Scott, during which he will answers questions that were submitted by the community. On Day 2 I will be interviewing @SteemMonsters co-founder, @yabapmatt, who will be sharing the insights he's gained from working inside many startups (and founding one of his own) in order to help blockchain entrepreneurs and developers gain a deeper understanding of the challenges they will face as they build their business.
