Filtir

Filtir is an AI tool designed to help fix “hallucinations” in text generated by ChatGPT, a powerful writing tool. The tool uses a state-of-the-art biologically large language model to determine whether the factual claim is true or not, based on publicly accessible evidence available on the open internet. If appropriate evidence cannot be found, the claim is considered hallucinated and removed from the edited output. Filtir also provides links to the evidence used to support its claims, however users are still advised to check it themselves if needed.

The team behind Filtir consists of Samuel Albanie, Vlad Bogolin, Anthony Buonomo and Ioana Croitoru. Their website includes an About section as well as Frequently Asked Questions and links to their Privacy Policy, API Data Policy and Terms of Use. Furthermore, Filtir takes user privacy seriously and does not sell or share data with third parties for marketing purposes.

Using Filtir may result in slower performance due to its reliance on a language model that needs to be active at all times. Despite this small downside, Filtir can still prove useful for writers who want to produce accurate texts without “hallucinations” while being assured that their privacy will remain intact.