Twitter Starts Trial Operation of “Birdwatch” to Prevent False Information by Adding Notes from the Community | GLM

Twitter has launched a “community-based approach to misinformation.”

This “Birdwatch” project is a master of new feature discovery. J It first came to light last fall by (Jane Manchun Wong).Currently Twitter Birdwatch website The pilot version has started operation in the United States through.

The purpose is of Keith Coleman, Vice President of Products on Twitter. Blog article As explained in, it’s about creating a misinformation prevention system that goes beyond the “labels” that the company has already applied to controversial and potentially misleading tweets. According to Coleman, the tweets that are labeled are limited to “a situation where something breaks our rules or gets widespread public attention.”

Birdwatch’s approach “will broaden the scope of voices addressing this issue,” Coleman wrote. It brings more perspectives to these issues and goes beyond just the question “Is this tweet true?” But the effort may deprive Twitter of the enthusiasm for users to check and rate individual content.

Users can sign up for Birdwatch’s site to flag potentially misleading tweets, add context with notes, and rate notes written by other contributors. be able to. These notes are currently only visible on the Birdwatch site, but the company’s goal seems to be to make them available on the main Twitter account.

“We believe this approach has the potential to respond quickly to the spread of misleading information by adding a context that people can trust and value,” Coleman said. .. “Ultimately, we aim to allow Twitter audiences around the world to view their notes directly on tweets so that they can get the consent of a wide variety of contributors.”

Given the potential for many discussions and counter-arguments from tweets, it’s interesting to see how Twitter can avoid confusion and mess, or prevent concentrating on some counter-arguments. .. According to the company, Birdwatch ranks content based on an algorithmic “evaluation and consensus system,” and the code is public (all notes posted to Birdwatch are available for download). You can read more about the early ranking system here.

“We know that there are many challenges in building such a community-driven system. To prevent manipulation by specific groups, to avoid being dominated by simple majority opinions, We will focus on these things during the pilot period, until we ensure that the distribution of contributors is not biased, “Coleman said.