I've played with the readability lib by Mozilla before to extract articles from web pages and then feed that to the summarizers we had years ago but the result

I've played with the readability lib by Mozilla before to extract articles from web pages and then feed that to the summarizers we had years ago but the result wasn't great. This is indeed different now, interesting! Another thing that would be great is to use that to identify the news that really matter and mute the rest. I could imagine a prompt containing something like "As a person that likes A, B, C, D, and F, does the following has a potential to be interesting to me: ..."