Bouncer: Block "crypto", "rage politics", and more from your X feed using AI

2 days ago (github.com)

Why, when you could just not use X though?

Don't you have to pay Twitter to get programmatic access to a X feed? [1] The documentation mentions using the "Twitter adapter", which uses a paid API.[1] Using an unofficial client has been a TOS violation for many years now, since Twitter killed off TweetDeck.[2]

I used to have an ad filter for Twitter, but gave it up a decade ago when they changed the TOS.

[1] https://docs.x.com/overview

[2] https://cdn.cms-twdigitalassets.com/content/dam/legal-twitte...

Hasn't everybody already figured out how to do this with a Mute Word list? The topics and names that trend on Twitter are not exactly a mystery. Every post of that nature follows a predictable pattern of key phrases/terms plus stupid emojis in order to "go viral", and can be filtered out easily.

  • I have spent a long time maintaining mute word lists but it's never more than about 80% effective because obviously there's a lot more to content than mere keywords. Tons of false positives (trump as a verb) and negatives ("he's blocking the strait!!")

  • It's often very frustrating when you care about something that shares a name with the ticker symbol for a popular shitcoin

  • You could also just be careful who you follow and constantly curate.

    • That doesn't work, regular people who aren't the audience for ragebait don't realize that quoting ragebait tweets with a sardonic reply is a positive signal for the algorithm to circulate that tweet even further. Mute is the only way to go.

There would be absolutely nothing of substance left on X without crypto and rage politics

  • America and Japan have been discussing BBQ for about a week and. And the Japanese have just discovered ranch dressing. It's hilarious to watch.

    Russians recently joined in and, they're fuckin' hilarious in a way that only Russians could be: https://imgur.com/GaTnQk7.jpg

    The people who hate Musk/X always out themselves by flagrantly lying about it/him.

I feel like regex and curated blocklists would get you pretty far before needing an LLM to continuously read your feed. I'm wondering how successful the local options are, because sending your social media feed to an API that is also being used to serve you low quality posts your blocking is a pretty depressing ouroboros.

Is there a tool to undo the extra weight added to paying subscribers? Analysis shows premium subscribers end up with 10x as much reach on average than people not paying.

Pay2Play was toxic enough on gaming, why would we want it in our social media?

You can also get excellent results with User Scripts that connect to a local language model.

1. Run LM Studio (download the Gemma 4 model), which has a 'local server' w/ API.

2. Use a more powerful LLM to write User Scripts (Greasemonkey, etc) to do whatever you want on any website you go to. Instruct it to connect to LM Studio.

3. Classify and highlight posts / comments based on any criteria that suits you. Summarize, delete from the DOM, etc., Just have fun.

I would also recommend Control Panel for Twitter, easy to install on PC, and you can also have it for your phone if you use twitter via the Firefox mobile browser.

If all you can see is the following tab, then any ragebait that gets in your way is much more actionable, simply unfollow or mute whoever got it on your feed.

It should also click "see less often" on every detected bait post. Heals the algo really well if you do that persistently

  • The fact that you use the word 'persistently' seems to somewhat (or completely) undermine what your point.

  • On Facebook, at least, the click seems to outweigh the feedback.

    I say "not interested" to a reel and get more just like it.

    • I get the ads about Warren Buffet’s (or other money celebrities) investment group or whatever. They are usually WhatsApp based pump and dump schemes for Chinese stocks.

      Facebook somehow can’t detect these obvious scams, but somehow they have no problem pushing them to me after I looked into it when a fried almost got taken.

    • Exactly my experience a few years ago (it not working is directly related to how little I use Facebook today). You might stop getting stuff from that specific page or account or whatever but you certainly continue to get related stuff.

TIL: there are still people out there who find value in x even though when I log in my feed is just hatred and false info.

Xtwitter’s own mute words feature is very good . And mute words supports TTL. LLM will have precision / recall issues too – no filtering system will be perfect.

Cleaning up 90% for free is better than burning tons of tokens / GPU / battery to clean 95% (and suffer from false positives).

I'd like to just quit twitter, but unfortunately the other places devoted to discussing some of the hobbies I go to twitter for, are much more toxic (Reddit, 4chan etc). Simply being able to filter out everything unrelated to the hobbies I'm there for would be sufficient.

  • Could you expand on this a bit? I'm deeply interested in how one could make a not terrible social network and I'm curious what gets people so unable to walk away

    • I'm into niche VTubers and other anime stuff. Twitter is pretty much the primary platform for keeping up with Vtubers (besides wherever they stream).

      The other platforms don't let you interact with or keep up with the vtubers directly, and often involve just mindlessly repeating the same joke, or they go the opposite route and take things so seriously that all discussion is stifled.

      With other anime stuff, Twitter is the easiest way to keep up with the Japanese side of things.

      Lately I've learned about tildes and while I haven't looked around much, it has me wondering if maybe invite-only forums with low barriers to entry (and low barriers to being banned) are the way to go.

I have an easier solution: just don’t use X.

It does not have anything worth following, and useful content is minimal.

I find that using Control Panel for Twitter (not affiliated, just a happy customer) to see only the Following tab in reverse chronological order makes X tolerable. There is no benefit to For You.

What makes X usable for me is the keyword blocking. You can also block posts that contain certain Emojis, including in user profile. This _really_ helps get rid of a lot of crap.

I'm amused at thinking of the other effects this can be used for, rebrand it as a tool like that copilot recall and point it with child privacy in mind for the general internet.

or you know, require it for internet/computer usage for a very dim futuristic outlook.

I don't think that automated filtering on conditionals like "rage politics" is a good idea. At best, you're going to end up with a confusing feed that contains reactions to the outrage without the actual outrage that's driving them; at worst, you're going to end up systematically misinformed on political topics that people find outrageous.

  • "you're going to end up systematically misinformed on political topics that people find outrageous."

    If you spend too much time on X, that's a given. The problem is that informed, nuanced, and factual takes don't drive clicks and are hard to fit in 140 characters. Long-form Youtube is a much better place to find those types of takes anyway. Generally, the shorter the content, the worse the take.

  • > you're going to end up systematically misinformed on political topics that people find outrageous.

    That sounds... fine?

    • I would emphasize misinformed, not uninformed. If Policy X has 30% of people politely supportive, 20% of people politely opposed, and 50% of people incandescently furious about it, you're going to mistakenly think it has majority support.

      5 replies →

I get the idea but honestly asking: if you filter out stuff like this will you end up with a completely blank feed on x? To me it kind of just seems like we're all going to need to curate our own RSS feeds in the future. eg: real people who are insightful, rather than rely on any kind of algorithm.

  • No, there's approximately just as much technical and interesting content on Twitter as there used to be. Lots of people left, lots of different people joined.

    It's just that this content is outnumbered some 100,000:1 now instead of the mere 1000:1 it used to be (ratios made up, but directionally correct.)

    From my point of view, HN is trending in that same direction. It's just that the ratios aren't nearly as dramatic.

    • It's the ratio that counts the most. You seem to be implying TwiX is getting an increasingly bad ratio. That would imply, to me, an increasingly limited lifespan for encouraging quality.

Adding to the chorus: if you need to apply a solution like this, it's probably time to walk away from the platform. (Well, the right time to walk away would have been years ago, but...)

  • All remotely popular online public spaces are completely infiltrated by bots/propagandists/trolls/morons/etc. If you could successfully filter that type of content out you'd end up with a much larger pool of valid/authentic content to access than if you abandoned the space altogether and switched to some very obscure/niche space that's yet to be manipulated.

    • You can already follow who you want on Twitter. The thing is, bots etc take their toll even on the good users.

    • Bluesky has a default feed that is just the posts/reposts of the people who you choose to follow, in reverse chronological order.

      No need for an algorithm to decide what is worth seeing.

      3 replies →

  • We have a solution like this for HN, but people don't use it: It's the "hide" button, and it's right next to the "flag" button. Yet, when users see content they don't like, instead of just hiding it, to block it for themselves, they often choose to flag it so that they can block others from seeing it too.

    I'd welcome per-user curation tools like OP's which don't affect the content for the rest of us.

  • I was actually thinking of making a similar app for hacker news comments. Should we all quit hacker news too?

    • HN is my top candidate for a solution like this, too. Because there's a ton of high quality content here, increasingly buried beneath a small number of sentiments and topics I don't care to see rehashed constantly.

      2 replies →

    • HN doesn't need it. I'll read this site, not gonna bother with Twitter or Reddit though.

  • Network effects are stronger than we are. People are there because people are there.

    • And when you are not there you are not there. We are way too obsessed with missing a thing. May it be a popular figure or someone we know in person. The reality is that it's actually not too bad to miss things and most information still gets through. Especially the one that's important. You might even miss out on a lot of crap that is filtered out when it gets to you.

      I am happy on my personal Mastodon instance and occasional visits to HN. You might be too if you allow yourself to be.

      2 replies →

    • yea but which people ;) unless you want to in that in-group, crypto, rage and all, better off without it

    • I know a bunch of people and companies who happily dumped the twitter cesspool. It has to be > 50% scammers and ragebots at this point.

  • You need to curate your algorithm. Took me 10 years before I started blocking aggressively and now my feed is amazing with 90% bangers. Twitter is by far the best product in this space. Every other platform is 2+ weeks behind. Twitter is where the news breaks.

    • I had a well curated feed too (even used word filters) and yet I felt compelled to pack up and walk away. It was simply not enough.

      The negative effect the various drivel had on me was nonlinear. Even if 99% of posts were fine, if that 1% was seriously upsetting, it just ruined the whole thing.

This would be great for Reddit - the king of rage politics.

  • The problem with Reddit is different. Poor quality human moderation is the problem there. Basically who has 10 hours a day to read Reddit? Answer, terminally online bubble people who have no business moderating other's posts. Maybe if the LLM could completely bypass the moderators then it could work though.

    • When I tried Reddit a while back, that problem showed up even with no moderation action being taken. I guess cause an obsessed person will use the site like 1000x more than a regular person, they end up being the "majority." The voting system also encourages bad behavior.

Oh thank God I can still give Elon Musk more money and power but don't have to read about politics

What I don't understand how difficult it seems to be for some people to simply ignore topics or people they don't like. If an algorithmic feed keeps presenting you with certain topics, it's largely because you're engaging with them. Isn't that on you?

I don't use Twitter but I use Tiktok and you know what I do when I see something I'm not interested in? I scroll up. If it's someone who never has anything interesting to say, I just block them. And I never think about them ever again.

I rarely see anything about crypto. I don't even think about it really. Go back ~4 years and everything on HN was about crypto this and blockchain that but that's how it goes. There are fads and, more importantly, there are people just trying to get their bag with their latest "acquire me please" startups. Actually, crypto just had a bunch of straight rug pulls too. And then there was NFTs...

Anyway, I've worked for my Tiktok fyp. It's a constant moving target for the platform too, like these bot accounts that somehow get to 10K followers and then appear on your fyp with audio over a movie or TV show to get around copyright detection. I honestly don't know how they haven't solved that problem yet.

All these platforms, particularly Twitter, put their thumbs on the scales about what gets distribution but for any platform with a block feature, this seems like a "you" problem if your feed isn't what you want.

Also, "rage politics" in general just means "things I disagree with" whenever anyone talks about what they see on any social media platform.

Block and move on.

  • When I was on Mastodon, I followed a bunch of people for their tech expertise: FOSS, security, networking, Linux.

    They also posted about other topics where I wasn't interested in their commentary. Even when I agreed, I still didn't want to see it, because I went to Mastodon solely for tech. I had other sources for other topics.

    So I added a bunch of filters to exclude those posts. It worked well!