In terms of research, Wikipedia, though it provides a very preliminary level of data, is indeed the first step. If you have ever gone through Wikipedia, you already know how pathetic the experience is, and how easily the information can be changed. You even question the truthfulness of the data and facts.
Now, there are thousands of volunteers who not only check the credibility of information but also go through the sources cited on the website. Now, in this large equation, comes AI. According to a new study, AI can now clean up the mess, including incomplete and inaccurate references on the website.
SIDE, a neural network created by the London-based company Samaya AI, is designed to streamline the analysis of sources referenced in Wikipedia articles. It helps ensure that these sources align with the article’s claims and can even provide alternative reference suggestions when needed.
The parent company, Samaya AI, was founded by Maithra Raghu and Fabio Petroni. Maithra is also one of the brightest minds in the tech industry. With her work experience at Google and her degree from Stanford University, she is poised to do something great for the future..
In terms of the training of the AI, SIDE has been trained using a database of reliable references found on the website, which often receives recognition from editors and moderators.
Now, if you talk about user experience, 21% were happy with the AI references, 10% of the users opted for what was previously available on the website. And the rest declared not to state any opinion.
