All US users have been added to the Birdwatch community moderation program.
It is a big step for Birdwatch and marks a step up in the platform's efforts to reduce the spread of misinformation. The most common topics being fact-checked are already covered by Twitter's misinformation policies, raising new questions as to the impact of the program.
Birdwatch wants to decentralize the process of fact-checking misinformation by putting power in the hands of the community of users. There is a huge range of topics, from trivial and easily debunked rumors to complex claims that depend on fundamental uncertainties in the scientific process.
It can talk to the internet's random curiosities.
The easier decisions have been the focus of public statements by the executives. In a call with reporters last month, Coleman suggested that the strength of Birdwatch was in addressing statements that were not covered byTwitter's misinformation policies or weren't serious enough to be assigned in-house fact checking resources. Coleman said that it could speak to the internet's random curiosities. Is there a big void in the sky? Is this bat large enough for a human?
The Birdwatch data was downloaded by us. There were 37,741 notes in this dataset.
We used the natural language tool kit library to extract the most important words from the notes.
We excluded words that were frequently used in the process of constructing a fact check, such as "but", "there", "which", and "about." We reduced the number of parentheses to their singular form, so that cars would be counted as 'car'.
A good overview of topics that are commonly addressed or have context added to them is given by the processed data.
Many Birdwatch users are trying to tackle more serious misinformation issues on the platform and are trying to overlap with the policies of the social network. COVID-related topics are the most common subjects addressed in Birdwatch notes. Many of the accounts that posted the annotated accounts have been suspended, suggesting that the internal review process is catching content violations.
The regularly updated dataset of all Birdwatch notes is free to download from the project's website. The data spanned from January 22nd, 2021. to September 20th, 2022. We can gain insight into the major topics of Birdwatch notes with the help of Computational tools.
The data shows that Birdwatch users spend a lot of time on the internet. The related term "vaccine" is at number three on the list and it is shown that "COVID" is the most popular subject term.
As public understanding of the Pandemic changes, the types of claims being fact-checked evolve. There are false narratives that Dr. Anthony Fauci had a role in the creation of the novel coronaviruses or the safety of vaccines.
ivermectin and hydroxychloroquine are some of the treatments addressed in Birdwatch notes.
There are some COVID myths that are easy to fact-check, such as the idea that the virus was a hoax, or that it gets spread by 5G towers.
Birdwatch user tried to add context to an argument over one vaccine brand's effectiveness at preventing hospitalization vs preventing any infections. New Jersey Governor Phil Murphy said trial data for the Johnson & Johnson vaccine showed "COMPLETE protection against hospitalization and death" and provoked an angry response from a statistician who linked the trial data to the vaccine's effectiveness.
Birdwatch noted that the author was confusing the reported efficacy of preventing hospitalization and death with the reported efficacy of preventing infections.
A Birdwatch user tried to fact-check a claim widely reported by mainstream news outlets, using a post on a website as a citation. A post on ThePrepared.com argued that the claim that the omicron variant made up 73 percent of new infections may have been a mistake. Without confirmation from a more reliable source, it is difficult to know if the post helped or hurt the situation.
Birdwatch users said that these are some of the most difficult to deal with. If users fill in a survey when creating a note, they can rate the claims on a scale of 0 to 100. It's clear that accurate, accessible communication of scientific findings is difficult, but public health outcomes depend on accurate health advice. According to experts, platforms need strong, clear and coordinated standards for addressing misinformation, and community-driven moderation is unlikely to meet this bar.
Covid is a main topic, but it is not the only one.
The word "earthquake" and "prediction" are highly ranked due to a large number of identically worded notes that were attached to the accounts that claimed to be able to predict earthquakes.
There is no proof that earthquakes can be predicted. One of the worst offenders is theQuakeprediction account, which has 48K followers, and posts a steady stream of earthquakes in California. A Birdwatch user attached a warning note to more than 1,300 posts from earthquake prediction accounts, each time linking to a debunk from the US Geological Survey explaining that scientists have never predicted an earthquake
It's not clear why the user focused on earthquakes, but the end result is a human reviewer ironically behaving more like automated fact-checking software.
Efforts to contest the results of the 2020 election are clearly shown in the data.
The terms "Trump," "election," and "Biden" can be found further down the list. The claims that Donald Trump won the election or that Joe Biden lost are addressed in many notes. The overwhelming amount of evidence against widespread electoral fraud makes it easy to fact-check claims.
Joe Biden was elected. Wendy Rogers, an Arizona state Senator who is linked to the white nationalist movement, made a false claim about fraud occurring in highly populated areas on her account.
It is almost impossible to commit mail-in voting fraud, and there is no evidence the election results of 2020 are the result of fraudulent activity.
The election was not rigged according to another user. Trump didn't win. As with many other cases, the original tweets cannot be reviewed, and a message that the account has been suspended can also be found on the blank page.
Birdwatch users have annotated many of the results of the election, but the self-evaluation surveys rate these as less challenging to address, given the amount of evidence supporting Biden's victory.
The large number of suspended accounts makes it easy for the human moderation team to flag and remove the same content.
Data from the Birdwatch program shows a strong community of volunteer fact-checkers. The evidence shows that there is a lot of overlap between the type of tweets these volunteers are addressing and the content that is already covered under the existing misinformation policies of the company. Birdwatch should be added to existing fact-checking initiatives rather than any kind of replacement for misinformation controls, according to the social networking site.
According to the company, people are less likely to agree with a potentially misleading message if they see a fact-check note attached to it. Many viewers are still being taken in by false information despite the promising finding.
A request for comment was not responded to by the company.
All Rights Reserved © 2024
