- 'nytimes', 'theverge', 'theguardian' or 'npr' (politics, pop news and US business);

- 'bbc' for all those reasons and their inaccurate reporting on tech and science for the sake of accessibility;

- 'femfosec' (can we not?); and

- 'krebsonsecurity' (for dramatising and clickbaiting fairly mundane stories).

I'm fortunate enough that I hardly have to deal with FAANG where I live and work, so I also block 'fb.com' and any Google brand TLDs since I have trusted circles that act as a human filter on the interesting parts. I also previously blocked '.dev' because it's unavailable from the networks I'm usually working on, and they were usually vapourware email farms or CV padding.

For titles:

- /sexual|lgbt|pedo|culture war/ (unfortunate combo, I know) hides most US politics;

- /america|united states|\busa\b|u\.s\.|senat(?:e|or)|biden|trump/ hides the rest;

- /cov(?:id|-2)|corona(?:virus)?/ isn't what I'm interested in;

- /\w+js/ hides the JS (framework|library) of the (week|month);

- /in mice/ to hide the deadend studies.

I've also started filtering titles beginning in /^how\b/, since these are almost always advertisements ('How to do X with $product') or political essays. /\(\d{4}\)/ also hides links to past articles, since the graverobbers are usually just trying to make a statement by invoking the wisdom of old. The hidden gems make their way to me through other channels.

I use this exact same filter on several other aggregators with similar success, so I don't think it's specifically an HN issue. For example: when applied to entertainment aggregators, I only get the entertainment news I came for. HN is my best place for a general feel of the industry's pulse, it just has some fluff in the way.