I was curious what would happen if you add many. Turns out that around 1000-3000 feeds the chaos starts to self-organize. I have 30 000+ subscriptions now and it's just wonderful.
Beyond the sweet spot you cant not filter out topics covered by everyone. You are aware Tucker interviewed Vladimir, there is Genocide in Gaza, Bidden has nothing interesting to say anymore. You just cant have tens of thousands of articles about David Bowie dying. You know that already.
And then.... when you've silenced the echo what remains is all the other things that happened in the world, the small, the unique, the rare, the interesting... the real www?
The funniest thing was businesinsider: They echo a lot but their titles are so descriptive they hit pretty much every filter I've created. That what remains is actually interesting. I think something like 0.01%! The next website with its "You will be angry when you find out about this!" articles must be extremely fascinating 100% of the time (this never happened) to avoid getting blacklisted.
I do quests like find feeds for press releases for all fortune 500 companies.
Best one thus far was to take a list of all countries and find 1-5 English news feeds for each. Turns out many have an English version of the national news that absolutely no one reads. Who is to say that if you have a few million people, nothing interesting happens there? Their official stance on events on the other side of the world is often so free from bias it is almost boring. A war becomes just a war without good or bad guys.
Whatever government websites publish is usually worth reading the headline of.
HN is a good website but the curated list of headlines on the front page requires the topic to appeal to the audience, the audience isn't as broad as other websites but it's a serious limitation compared to that what you find interesting.
If HN was to one day decided to force onto the front page each new book about eating bugs there is probably one guy out there excited about it. Then comes breeding bugs and then growing plants to feed to your bugs... Our guy would be thrilled? Sadly for him it cant happen. It's unthinkable.
Facebook is actively filtering out the things I want to read. lol?
Twitter is the mega echo chamber.
Reddit echo's and filters and is littered with low effort commentary. Subs are dominated by majority perspective.
Wikipedia is a trench war.
But what were we thinking? Why should other people be held responsible for what you read and write?
There is also the angle of programmers trying not to be political or activists. Over the years the sum of little crumbs of avoiding politics or activism adds up to enshitification.
When Musk fired so many at Twitter people talked about how difficult it is to get the posts onto everyones feed.
Meanwhile my crappy laptop and my crappy internet over my crappy wifi can deal with hundreds of thousands of feeds, millions.
I can read Nazi's, Maoists, Israeli's, religious fanatics, hackers, eugenicists without them making an effort towards making their thing more palatable to the masses. Non of them are making thousands of accounts in an effort to promote their views. There are no likes, views or upvotes to be purchased.
When their new website finds their way into my aggregator (how?) I can get rid of it in a single click.
It this point, I would probably invest some time into automating it even more. Download all articles, and save them into PDF. Create full text search over them. Use LLM for sentiment analysis...
I spend 3-10 minutes scrolling over at least 1/3 of 5000 headlines. Today I open 12 in tabs doing that, I've looked at 3 of those.
I got rid of most things, I only keep [pubDate, url, title] and only if the title doesn't contain any badwords and the pubDate isn't older than the oldest item.
Each item comes with a hours and minutes ago a translate to English button, the domain name, and a blacklist button.
While it does take opml most of my feed lists are just txt files with one url per line.
It loads one of those files then pulls x feeds per second and parses y per second where x and y are adjusted to the pending requests and pending parse jobs.
While you could certainly do many interesting things most of my experiments didn't survive my desire to keep it simple. I for example use to dump the results on a web page. I couldn't convince myself it was useful to the goal.
I had tried RSS years ago and my feed then was unwieldy enough I never checked it. This attempt is smaller and stuck