This is an attempt at creating a democratized, open-source website that does fact checking.
This project's API conforms to the OpenAPI standard and is automatically documented.
There are already some well-known fact-checkers out there, but none are perfect.
-
They are slow
Misinformation spreads like wildfire, and we can't always wait for the experts to discover it and write up a full rebuttal.
-
They aren't adaptive
What happens when a source is discovered to be untrustworthy? When an expert becomes discredited? When a scientific study is overturned? When fact judgements need to reevaluated, this should be discovered automatically. Modern technology can do far better than the word-of-mouth system that still seems to rule journalism and academia.
-
They are avoidable by people with fringe opinions
Existing fact checkers may actually do very little to change people's minds. Let's get these people involved in the conversation and introduce them to the tools for real skepticism by allowing them to provide context for their opinions.
-
They are easily written off as elitist
Linking somebody to Snopes is a great way to destroy a friendship. Fact-checkers should initiate conversations, not stifle them.
-
They are Anglocentric
Existing fact-checkers primarily provide content in English and focus heavily on US politics. The open source community at large can do better to provide information to a wider demographic.
-
They provide sources
Skeptics need to be allowed to investigate claims for themselves.
-
They are transparent
Without the guise of anonymity, fact-checkers require experts to stake their reputations on their fact judgements.
-
They rely on experts
We still need the guidance of experts to navigate nuanced and complex topics, and provide a summary for laypeople.
-
They aren't able to be manipulated
A 'democratic' approach to fact-checking must not itself be susceptible to propaganda and misinformation.
Imagine the following scenario: a high-profile politician tweets out a claim. A short while later, person A links the tweet on this website as a claim. Shortly after that, person B reads the claim and remembers that they saw a study that seemed to disprove this claim. So, they link that source to the claim as evidence and submit it as disproving the claim. This study is now publically linked as context to the claim, but remains flagged as unverified. A while after that, a user marked as an expert in this field finds the unverified evidence, and reads the study. While the study disputes the claim, the expert is not convinced it disproves the claim. The expert sends feedback on the evidence to person B with the revision that the evidence only disputes, not disproves the claim. Person B accepts these changes, and now the claim is marked as disputed by the verified evidence.
This project could take inspiration from some of the best of the web.
Wikipedia provides a democratic approach to editing that would be very apt here.
StackExchange organically attracts experts into conversations with laypeople on a very wide variety of topics.
Quora seems to do an even better job of organically attracting experts, but the quality of information here can vary wildly.