Dangers of Bias

As outlined in Redefining "Community-Powered," a strong ethical standard guiding the question formulation process from moderation to publishing may be essential in establishing Hypothesis as the destination for an unbiased, unfiltered accounting of what is happening in the world at any given time. Where contemporary publishers have created a business of editorializing reality in the name of attracting eyeballs, Hypothesis deconstructs the foundation from which the notion of subjective truth arises and rebuilds it with one that fosters objectivity—however idealistic. But that idealism has two edges, in that it can obfuscate the complexities inherent in the relationship between language and the way we interpret our surroundings while relying too wholly on a sort of Chomskian “common sense” interpretation of truth.

How many unique ways can a seemingly-simple yes-no question be posed? In what ways could each of those formulations impact a user’s likelihood of choosing a particular outcome? What, if any, language implies bias toward a particular outcome? In what ways can question language perpetuate stereotypes and political agendas? These are only a few of the considerations that must be investigated before a single question is asked and a single prediction is made on Hypothesis feeds.

We must also establish best practices for moderation of submitted Context and the discourse surrounding it, as well as considering the ways and reasons outside actors may leverage these input channels to influence popular opinion toward particular outcomes. In which scenarios could the belief in a particular outcome influence the outcome itself? Should Hypothesis avoid posing questions about events where this could happen, or is there a way to protect against it? With bias capable of creeping into almost every aspect of the platform, Hypothesis will remain committed to innovating tools for fostering a constructive, minimally-corruptible environment for pondering the future.