Media Sifter: How a Decentralized Community Can Truth-Hack Fake News on the Blockchain

Media Sifter uses the blockchain to create the evidence economy — restoring the link between facts and the news.

This article is a collaboration between Thomas Schrum Nicolet, a philosophy graduate student, and Cy Rus Co-founder and Content Lead at Media Sifter

Forget Bitcoin. Attention is the highest valued asset in the information age, giving rise to the attention economy that underpins the digital universe. The current incentives for publishers to seek as much attention as possible has disconnected news from facts. The result is the rise of ‘fake news’ and widespread distrust of the media. In this article we dissect the dynamics of the attention economy, its relation to the media industry and the threat it poses for modern democracy. It’s not all doom and gloom however. Thanks to blockchain technology and innovative ideas, a new breed of decentralized models for information exchange are emerging. At the end of this piece, we’ll present one of these, Media Sifter, a new breed of decentralized platform. Media Sifter is offered as an alternative to a media landscape based on attention, re-aligning financial incentives with the truth, and in turn helping to restore our trust in the news.

As Herbert Simon wrote in 1971, the enormous wealth of information entails the scarcity of what it consumes; the attention of its recipients. Hence, as Simon writes “a wealth of information creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it”. This is the background for the attention economy, the digital landscape we all live in.

Since we can only direct our attention to one place at a time, our attention is finite, and thus an increase in information also entails an increase in the value of attention. The digital era created an explosion in the amount of information we produce and consume daily, causing the value of attention to skyrocket. No one knows this better than tech companies, who have built entire businesses based on Metcalfe's Law, a principle which dictates that as more users are gathered to a service, its value to the users becomes exponentially higher. It holds that in platform economics, internet businesses must seek to gain our attention and indeed the attention of as many users as possible in order to build value.

However there is a dark side. The attention economy is a zero-sum game, when you allocate your attention to one place, it’s lost somewhere else, an opportunity cost which we are often unaware of. As technology gets increasingly better at getting our attention and keeping it through addictive patterns, they covertly hijack our agency. This is a problem in its own right to be dealt with, but it poses a more immediate threat to a special business of information: the media.

The media is incentivised to continuously optimize the amount of attention that their news can gather. However, in the current media landscape, how much attention a news story gets has become detached from how accurate the reporting actually is. Or said another way, the ability for news to grab attention is independent from the news actually being true. The attention economy then, has created the conditions for a media ecosystem that rewards sensationalist headlines, outlandish claims and the misrepresentation of events. Indeed nearly nothing gets more attention online than ‘fake news’.

A possible definition of ‘fake news’ could be information that consists of a mixture of true and false information, intended to appear like informative news, while really optimized to grab the highest amount of attention, in order to further a financial, political or otherwise personal gain.

This leads to what has been called post-factuality. But surely the onus is on us as media consumers to vote with our clicks, and choose factual stories? Well, there are many reasons why this fails to happen. Due to confirmation biases, we often pick news that fits our identity or current opinions. Given an abundance of information and limited attention to allocate, this tendency is strongly enforced. This is not new. But as publishers in the attention economy are incentivised to release news that gets the most attention, and news gets most attention when it fits narratives, it's in the publisher's interest to publish news that fits existing ideas.

Similarly with tech covertly hijacking our attention, we are mostly unaware of the disconnect between the veracity of news and its virality. A reason for this can be seen through the effects of groupthink and social proofs. When we are uncertain of something, we often look to our peers to see what they do or think. A desire for harmony or conformity leads to irrationally accepting beliefs or norms we otherwise have no justification for. There are often rather good reasons to do this too, formulated in the recently developed philosophical field of social epistemology.

From social epistemology, we know that, besides factual evidence, we might also rely on social evidence, which can be seen in the form of actions, thoughts, utterances or the messages of other people. We might very well be justified in believing something because of social evidence gathered from others, and the justification of this comes to hinge on how justified we are in trusting social evidence.

The downside of social proofs and groupthink comes into play when we misjudge or misinterpret the apparent behaviour of others. The issue is the failure to differentiate between people allocating their attention to a news source, and them subscribing to the claim it contains. That is, when we see a group of people clicking, liking or otherwise buying into a piece of information, we assume they believe the information to be true. Even if we are privately uncertain about the factuality of the information, we are likely to accept it as true if we believe others buy into it. This leads to a case of pluralistic ignorance, where a group of people can collectively believe something they privately reject.

We have already seen that the attention economy gives publishers incentives to propagate stories that receive as much as attention as possible, irrespective of truth. Due to social proofs, people are more likely to believe information that receives a lot of attention. Thus, we have a vicious cycle, where the likeliness of us believing information to be true is inherently linked to its ability to capture attention, while any connection to factuality is lost. In essence a strange paradox emerges; the likeliness of us believing information to be true becomes totally disconnected from the information actually being true.

Once we become either consciously or subconsciously aware of this disconnect between veracity and virality, we start to lose trust in the news. We enter a strange world where even truthful stories can be doubted and discredited, because they occur within an attention economy in which incentives are misaligned with truth. We also lose trust in social evidence, meaning we allocate less weight to what certain other people believe, say or do. This in turn leads to polarization, which leads to a downward spiral continuously enhancing these effects further.

These types of informational blunders can happen in all kinds of information environments but are especially problematic in a democratic context. Currently, the democratic ideal is coupled with the enlightenment assumption that if we are rational, we ought to base political deliberation and decisions on truthful information. The issue we are are facing is that while we might be individually rational, we can be collectively irrational, especially in contexts in which it is hard to ascertain the truthfulness of information.

Another idea from the enlightenment connected to democracy, is that if we just give people enough information, they will make good decisions. Yet today we see something different. We have seemingly endless information, and as it’s sifted through the attention economy, it leads to political discourse, deliberation and action to be based on information that tracks attention, rather than information that tracks the truth. This is an issue, as whatever political persuasion one might have, one would like it to be based on facts.

Contrary to what some might think, when people agree with ‘fake news’, they are not oblivious to the false information presented. What often happens is that due to confirmation bias, social proofs, groupthink and the collective irrationality described above, we cherry pick true information that we like and ignore what we dislike, or erroneously believe in false information due to irrational group behavior. We do tend to think that our beliefs are grounded in facts, while we might think those with opposing beliefs are victims of 'fake news'. In fact this might be the case for both parties in such a disagreement, and they might both be right.

What we see is somewhat the opposite of the enlightenment ideals: we have more information than ever, which should in theory be positive in regards to political deliberation and decision-making, given it can be correctly curated. Instead, this abundance of information makes us even more irrational, through polarization and echo chamber effects.

The missing piece, hinted at in the paragraph above, is the seeming impossibility of curating the enormous amount of information in a digital media world. Moreover, by whom should it be curated, especially when we find it hard to trust anyone as a gatekeeper, and particularly for news which might have real democratic consequences?

Media Sifter, the new fact-based news platform on the blockchain, offers a clear answer to this question: the curator should be Everyone.

In this White Paper, addressing the issues described above, Media Sifter seeks to offer more than just a new context for news consumption, it looks to fundamentally change the rules of the media game. Firstly it aggregates the news, making it possible to sift through some of the noise generated by endless information being thrown at you, the reader. By placing stories on the same topic, but with differing perspectives, side by side, you are confronted with opinions and points of view that are unfamiliar, in a context that is inherently unbiased. The Alpha version of the platform was recently made available to the public.

The core solution Media Sifter brings however, is the evidence economy; a new media ecosystem which rewards and incentivises facts. The team behind Media Sifter has worked at length with behavioural scientists and game theorists to develop the SIFT protocol, an open source, stand-alone blockchain protocol which aids in sourcing and validating evidence from the global crowd.

Within the Sift protocol are two types of digital systems, the first of which is The SFT Token which provides access to the evidence economy, and rewards investigators who contribute validated evidence to the platform. The second is The CRD score an on-chain immutable user credibility score, which rewards persistent investigators and other positive user-engagement. These work in harmony to tackle the issues inherent to the attention economy and to reinstall a link between facts and incentives.

For instance, the majority of financial rewards available on the platform are allocated to those who provide factual evidence to the aggregated news stories. This occurs when a Reader highlights a sentence within a news article that they want to be fact-checked. The reader places a ‘bounty’ of SFT tokens on the statement at hand. This action attracts further readers to up the ante, adding more SFT tokens to the bounty. Once the bounty reaches a lucrative enough level, Investigators are incentivised to go out and bring 1st, 2nd or 3rd hand evidence to either validate or invalidate the claim made in the article.

However, the immediate question is, how do you decide what is valid or invalid and indeed who decides what counts as true or false? For Media Sifter, the answer is the community. When a piece of evidence is provided on the platform to support or refute a claim in a news story, the evidence goes through a review process. Using a mixture of random selection and the CRD score, a panel of reviewers is selected. Their job is to validate whether the evidence provided responds to the initiated claim. This deliberation is done independently, with no collusion possible.

In these circumstances, if each person in the panel, working independently has a greater than 50% chance of selecting the right answer, the outcome of a consensus will converge on the correct answer. This model replicates the famous Condorcet Jury theorem, the results of which have proven highly difficult to demonstrate in practice. The theorem states that a jury of voters with any probability above 0.5 of choosing the correct answer, will always be better than a single authority, even where the authority on their own has a higher probabilistic chance of determining the correct response. Now thanks to decentralized technology we are able to make this theory a reality.

Interestingly, this turns platform mechanics on their head. While today it is highly lucrative to indulge in noise-trading, the evidence economy should in theory make noise-trading economically pointless, as you are rewarded for validated facts and evidence, not the amount of attention an article receives. Voters should converge on the truth, as on a transparent, decentralized platform they are aware of these mechanics, and being aware of others converging on the truth makes one more likely to base their vote on what one believe is true. This is especially valid since any false information is almost certain to be rejected by the consensus.

Another pressing question is whether or not reviewers will be affected by the same social proofs described earlier. In normal settings which generate pluralistic ignorance, people observe others before making a decision themselves, or at least have some idea about how others might act (think opinion polls and elections). On the Media Sifter platform however, people are asked to vote on an evidence claim before they know what the choice of the majority is and indeed without knowledge of who else is voting on the topic.

Reviewers on the platform receive financial rewards to perform this service, these rewards are provided in any eventuality and not contingent on the outcome. This might appear counter-intuitive, but this 2009 study from Yale, seems to show that people who are conscious of money typically strive to be more self-sufficient than those where money is not an issue. This would mean that the voting review process proposed by Media Sifter would incorporate people’s individual rationality, rather than leading to irrational group behaviour.

For Media Sifter, the decentralized technology of blockchain is vital in achieving the above behavioural mechanics. Built on the Ethereum blockchain network, almost instant transactions of all types can be achieved at a relatively low cost. Perhaps even more crucially in the context of today’s media ecosystem of centralized platforms and attention economics, blockchain allows for decentralized solutions, removing intermediaries and any necessity for trusted third parties. The result is a trustless system, removing the risk of outside interests affecting decision making, and a tamper proof network that is resilient to censorship and malicious attacks.

Through a token model, it allows for light micro-subscriptions on a news platform where readers ‘Pay As They Follow’, directly supporting the authors and publishers that they care about. But wait, nobody pays for the news anymore right? Well, while the advent of the digital era ushered in a decline in news subscriptions, more recently there has been a surge in support through online subscriptions for platforms promising quality journalism. We can see the feasibility of this from cases like the Guardian, which recently announced that they now get more revenue from supporters than from advertisements for the first time since they went online

The Media Sifter model serves to create an incentive structure that enables a media environment that tracks truth rather than attention. Trying to realign incentives with desired desires behaviour offers a realistic and compelling alternative to the impossible task of trying to getting people to pay centralized news publishers to produce factual news. In fact, if the incentives in the attention economy persist, it might only be a matter of time before even publishers succeeding in gaining digital subscriptions resort to lucrative attention seeking tactics in order to recruit new subscribers to their growing user bases.

Inhere lies the crux of the matter: hoping for corporations and online media platforms to stop profit maximizing will not create the conditions required for a truthful news environment. The underlying economic incentives must be re-aligned, such that qualities like evidence, facts and investigation are directly linked to economic success. Blockchain provides us an ability to rewrite the de-facto advertising and attention based business model, and restore much needed trust in the news environment. Through Media Sifter we have an opportunity to not simply create a fact-based news platform, but to change the rules of the media game.

Thanks for reading. Please give us a clap and share if you believe in our mission to decentralize influence and distribute truth —

— Join the Community changing the rules of the media game:

• Jump on the Media Sifter Slack or Telegram channel
• Read weekly articles on the Media Sifter blog
• Follow Media Sifter on Reddit and Twitter
• Subscribe to the newsletter
• Join the team and help build a fact-based future angel.co/media-sifter

For more information, visit MediaSifter.co