I really like the idea behind the brutalist report [1], mostly related to their claim of "headlines delivered without bullshit".
But I wanted even less interaction with most websites, as most of the headlines are related to something I am just mildly interested on, to the point of wanting to know more, but not interested enough to the point I feel like committing to reading the whole thing before knowing a bit more what it is about.
Anyway, I build this thing that parses the RSS feeds and do all summaries locally through ollama or whatever other OpenAI API compatible llm platform you have (e.g. LM Studio), and auto updates automatically twice a day through GitHub action scheduler
I really like the idea behind the brutalist report [1], mostly related to their claim of "headlines delivered without bullshit".
But I wanted even less interaction with most websites, as most of the headlines are related to something I am just mildly interested on, to the point of wanting to know more, but not interested enough to the point I feel like committing to reading the whole thing before knowing a bit more what it is about.
Anyway, I build this thing that parses the RSS feeds and do all summaries locally through ollama or whatever other OpenAI API compatible llm platform you have (e.g. LM Studio), and auto updates automatically twice a day through GitHub action scheduler
And it is open sourced at https://github.com/felipesabino/rss
[1] https://brutalist.report/