Citation allows you to confirm the information in Wikipedia articles, but it can be used in bad faith. For example, the article about Joe Hipp, a member of the Blackfeet tribe describes how Hipp was the first Indian boxer to compete for the WBA world heavyweight title. The article links to what appears to be a relevant web page – but it doesn’t mention Hipp or boxing.
Meta wants to use artificial intelligence to prevent authors from committing such fraud. The new development will be able to efficiently analyze information related to a quote and compare it to the source. At the same time, the meanings of the texts will be analyzed and compared.
In addition to being able to detect fake citations, the tool can suggest better sources to corroborate the information in the article. He can find the links that best illustrate a particular point. Theoretically, it can be not only text, but also multimedia materials. For example, a YouTube documentary that best covers the topic.
So far, the tool is a successful proof of concept that this is possible, but not yet usable.
“It’s not really usable at the moment. In order for this to be usable, you need to have a fresh index that indexes much more data than what we currently have. It needs to be constantly updated, with new information coming every day,” says a leading manager from the Meta AI team, who works with the project.