I'm really impressed with the quality of the output- backing it with search results seems to do a lot of good relative to ChatGPT's straight-up hallucinations.

How does it decide what search queries to run and which results to use? Any chance you could append the top sources to the output, maybe even highlight which claims are based on which sources?

Following that thread of thought, I'd love to be able to filter out commercial websites and blogspam as sources, or limit it to scientific publications and .edu/gov domains.

Nonetheless, this is a damn promising line of R&D.

Sorry, I didn't know this. A friend of mine (who actually gifted me Kagi a week or so ago) sent me the link yesterday. Very cool web app and it was extremely fast for me last night.