AutoVerbia Blog

 Explaining to the AI It's taken some work to do the extra explaining to the LLM on how to interpret the search phrase.  Though for the user of the search engine, it might seem simple.  The search engine can take almost anything you ask it, or even short phrases, and optimize them into the best phrasing for this specific engine.  The engine I'm using is a metasearch engine that searches other search engines in turn (SearXNG).  But it does have its own syntax that may or may not work with certain tokens.  The search engine components I used were open-source, so designing visual components was easy.  Explaining to the AI API behind the scenes how to formulate searches has taken many revisions, both as I try to speed up the search process and make the search engine more accurate.  Please leave a comment if you see any unusual behavior from the search engine; that way, I can improve the instructions in the future.  

Created A New Search Engine I got tired of search engines that provide an AI overview but don't help you find any results.    So I decided: why not integrate generative AI with the search engine?  What I came up with is AutoVerbia .  AutoVerbia   uses the assistance of AI to find websites on the internet.  It combines an AI API and a meta-search engine to produce superior results.  AutoVerbia   is taken from the psychology term autoverbalization, another word for self-talk.  This is as if to say, why doesn't AI just search for what is on your mind?  The search engine works, and rather than sifting through irrelevant results or guessing if the generative AI is right, you can find what you're actually trying to search for instead.  This search engine is a social experiment, and though it's one of the first engines to work with generative AI directly, I hope it's not the last.