You can't fight fake news using black lists.

The discussion around the "fake news" problem is so heated that we fail to see the real problem behind. If we want to differentiate among fake and real, we need to understand first how our minds differentiate among reality and fiction. As shocking as it might seem, the truth is that all news is fake in the strict sense of the term. This is why "fighting fake news" cannot be done using black-lists, it would be very easy to blacklist legitimate news as fake based only on the different point of view of the censor. In the words of the Romanian journalist Teodor Tita ”fake news is a fake problem". The solution may be in a positive approach, specifically in an algorithm able to rate credibility for the sources based on their entire internet history or a big portion of it.

What is reality?

First of all, we define something as real only in two situations: 1. we experiment that something with our senses, or 2. We have overwhelmingly credible testimony about that something.

In today’s world others testimony is the main cognitive instrument. Almost everything we learn about the world, news included, are other’’s people testimony. The problem is that the media ecosystem before internet had built in systems to preserve credibility. So that we used to simply equate news from credible sources as "reality".

How was credibility built?

Consider how news credibility was building before the Internet: you had a relatively small number of press institutions (TV, newspapers), in a highly competitive market, with huge audiences, reporting about "reality" with a high incentive not to lie about it because this would damage their reputation and reduce their audience to the point their competition will annihilate them overnight.

So we knew that for an important press institution to lie came with a very high, even fatal cost. But we also knew that while not lying, every voice describing an event may describe only a part of it and will use words to suggest a certain view on the subject. And this is why we always used to check other sources too.

In the end we have operated for decades with two assumptions:

  • lying is an unacceptable cost for any news organizations
  • the reality is always a sum of different accounts of a fact and never only one.

Internet came with no built in credibility systems

Internet change the news market in a number of ways:

  • There is a myriad of news providers, not only a few, and most of them are not press organizations.
  • News spreads by sharing among and between small communities, not by central dissemination to huge audiences
  • The chain of sharing which is engrained in the way Facebook works can easily make a single source look like multiple sources.

Internet in general and Facebook in particular work in a very different way from traditional media while we continue to use the old press assumptions about how we discern credibility. In fact an institution decided to manipulate reality finds that on the internet there is little cost to lying and is very easy to create seemingly different sources propagating the same message. But we still tend to believe that there is high cost to lying and multiple sources telling the same story make the story credible.

This confusion lies at the heart of the fake news problem.

Remember Yahoo (search engine)?

The chaos on the internet news front today resembles the chaos among the results of the first search engines. The ease with one could manipulate the position on the search results page basically turned those pages into commercial spam.

Google changed everything because it used for the first time an alghoritm to rank influence or "weight" which still is fundamentally impossible to manipulate — the now old and famous Page Rank. The idea was very simple at its core: influence increases with the number of other sites linking to it. This is a natural "vote"of the internet community which is very difficult to manipulate.

Steps towards a credibility ranking algorithm

However, credibility is different from Google's PR and is more of a Facebook than a Google problem. But we still need to find an alghoritm impervious to manipulation to rank sources credibility, more than building black lists which, however carefully curated, will be soon in themselves instruments of manipulation.

And the place to implement this algorithm is on Facebook because Facebook is the core network for news generation and sharing.

One simple way to approach this would be to use something like the already old Facebook "like": a way for users to vote credibility of the source and not a way to black list the fake news sources. And, like for the Google Page Rank, build an alghoritm which takes into account not only the number of votes but also their own quality of credibility acquired in time.

If building credibility takes a lot of input and time then it will be expensive to build fake credibility only to ruin it overnight. This can create a natural cost barrier (similar to Google's barrier to artificially increase PR) to fake news sources.

For sure there are enough bright minds at Facebook who can build on this positive approach and create a strong and feasible algorithm which can link into their news feed algorithms and make the fake news die naturally. No need to create small armies of censors and to make the problem worse.