Fake news, social networks and censorship
In recent weeks, following the outcome of the US elections, much has been written about the fake news on Facebook, and the need to control this. I’ve written about this in pieces such as “Facebook’s moment of truth”, “The inherent problem with technology”, “Trump and the dark times to come”, “The importance of developing critical thinking”, and “The dangerous chemistry of the internet”, some of which were cited by the media, in some cases associating my ideas with some form of censorship.
And before I go any further, I would like to reiterate that I am not in favor of censorship at all.
My comments about spreading false news and its possible influence on people were not intended as call for censorship on Facebook, but for the development of content qualification systems similar to those search engines like Google have been developing for years. The censorship of certain news or publications by Facebook would do little to nothing for the health of the web: action triggers reaction, and if the publications devoted to sensationalism and the systematic creation of false news were censored, it is very possible that Its popularity and diffusion, instead of retreating, would spread through other channels. Few things encourage conspiranoia and the desire to access information more than censorship.
What we have to do, as I have already written, is to try to promote the development of critical thinking, providing more tools to analyze information. I would recommend reading a recent great article by Hossein Derakhshan, the Iranian-Canadian blogger who spent six years in jail, published in the MIT Tech Review called “Social media is killing discourse because it’s too much like TV”. The web, as seen by a person who was kept in isolation for six years, has gone from being decentralized, based on text and conscious use, something analytical and providing links that broaden and deepen our reading, to something dominated by a relatively few channels that use — and abuse — images or videos that provide no additional information, and is consumed like television, and not subject to any scrutiny.
This not the result of any conspiracy, and may simply be a kind of evolutionary process typical of many societies, which tend to economize resources and to focus on simple solutions. The slope from simplification to simplicity is simple and slippery, and is one we have seen happen in different media: the existence and relative popularity of tabloid publications shows this. And of course, it is not about censoring and forbidding people to read The Sun, Bild, New York Post or other yellow media if they wish to do so. It is simply that people should know what they are reading.
Each year, Google is able to remove no less than one billion “pirate” search results. The technology exists and is certainly efficient, to the point that designing a system to do something similar with fake news would be, if not trivial, not really impossible. That said, we know that consumption of unlicensed content has not declined because Google has made it harder to find, but because companies like Netflix, Spotify, Apple Music or whoever have provided relatively cheap and easy alternatives.
Similarly, we need to start thinking about creating systems that are not dedicated to eliminating false news along the lines of censorship, but that instead label them while respecting the differences between humor, satire, slander, or invention. If an organization or individual does little else but produce fake news, that content should be tagged as such by systems that combine social news, machine learning and human supervision, blocking access to advertising mechanisms that function as and incentive (in the same way that companies like Google prevent access to their advertising from other types of content considered harmful, such as miracle products, pornography, etc.) and will be penalized by recommendation algorithms, to avoid the hall of mirrors effect. Is that censorship? Obviously, doesn’t treat all news the same, but then… should we really treat all news the same?
Any ideas?
(En español, aquí)
