There is a thing I generally believe about neural pathways. That the road we most travel is the one we'll keep going back too. And doom scrolling makes me walk down those paths. To the point that it impairs and guides my thinking even after I stop.

And I have stopped. I have had periods where I have blocked Reddit, HN, LinkedIn. I will find myself walking those paths regardless. And then, after a time, I start seeking out the doom again because it's comforting in it's discomfort.

It's probably the worst moment in recent history for this. The constant barrage of information whether it comes from AI, climate disasters, politics or any of the other myriad things that are occurring right now. It feels impossible to break from that cycle. I'm trying.

I want to be present and live in the moment. But it's hard to feel powerless. And Doomscrolling, while it absolutely doesn't bring power, brings a comforting illusion of it. I am aware of it, so I am at least a little in control.

Doomscroll doesn’t necessarily have to be about negative news. I don’t have any social medias but I’ve seen my friends feed, they are pure brainrot.

If you're someone who likes to do submit these things, please stop. We love HN as our little geek corner and refraining from political content would reduce the negative effects pointed out by the article.

So I'm building my own news aggregator. The value of this process has been surprising for me. I've had to interrogate what keeps me hooked. It wasn't actually the news. It turns out that's the easy itch to scratch, and it wasn't very itchy.

After some iterations and soul searching, I'm embarrassed to admit that part of what kept me doom scrolling was the rage. Arguing with strangers on the internet. Explaining to them why they're wrong. Laughing at silly comments by wrong people. In order of addictiveness, these are the things which keep me doom scrolling:

1. Rage. Editorials which oppose my value system. Articles which describe people who oppose my value system. Comments by people who oppose my value system. False/incorrect/bad comments.

2. Interesting discussion which extends beyond articles. This could be a list of interesting movies in a Reddit submission. New perspectives. Etc.

3. News. Especially around my niche hobbies. Often this news comes from niche sources which can be difficult to aggregate.

The third priority is technically easy to solve (well, mostly. I'm still working on it, but there is a clear path to success). I would like to completely eliminate the rage loop, so I'm omitting that entirely from the design. The challenge is therefore the second priority: discussion. How do I engage in productive discussion without the rage? They so effortlessly flow into each other in modern media - by design. It's so clear to see now.

I think back 15 years and I could have great discussions and debate with people on Reddit even when we vehemently disagreed with one another. Things remained respectful. Now it's more toxic than 4chan or Bluesky. So the quality of the community matters, but I think that is shaped by the algorithm and policies of the site. Reddit wants it to be a rage machine.

So how do I a) have interesting discussion, b) on niche subject news, c) without the rage? My current hypothesis to test is that I could use an LLM to intentionally hide all Reddit comments which could trigger rage. I might start with a browser extension. The major issue with this is that I form my own little echo chamber. All the comments and submissions conform to what makes me happy and comfortable. I'm not sure how healthy that is, either.

Further, Reddit is not designed around second-order algorithms. It's going to fight any attempts to subvert its rage algorithms. So building this on top of Reddit is like trying to drink sewage water by first sanitising it. I can do it, but should I do it? This raises more problems. We live in a world of monolithic platforms. The old internet is dead. I tried to spin up a Lemmy community and the only people who joined were actual communists who sent me death threats because I didn't agree with what Russia was doing to Ukraine.

I'm toying with the idea of creating an LLM-first/native distributed protocol which enables the communication of agents and their human users. The value proposition is aggregation: moving the discussion from closed ecosystems into whatever app/domain/interface the user feels like using. Communities appear and die and change as freely as users desire. Of course we already have a million social media protocols and the fundamental challenge is critical mass: if you don't have users, no one wants to use your platform/protocol.

Anyway, I've done enough rambling. This is a serious problem for me now so I'm going to keep chipping away at it. Thoughts and suggestions welcome.

Forums worked because people gently expressed their ideas/opinions. They collectively knew when to stop.

Now a kid can post something on the internet and hurt your feeling for days. Or a person who knows nothing about the subject can leave a remark about an event or famous people that they know nothing about.

My proposal is that your protocol should bring the concept of "social class". People only need to interact with people in their own class. It seems harsh. But that how society works.

You can't talk to the prime minister of your country. Even a big leader of a corporation in your daily life.

Old internet worked without any of this because it took effort to join a forum/thread/post in early days. Only people who are deliberate passionate about it join these kinds of "digital community".

Dead internet is real. But we still need to use technology to socialize. Make fences and don't waste time on public places where "uneducated" people can ruin your day.

I think that a carefully designed solution using LLMs could be the way, if it could correctly filter what I'd still want to read. I'm not sure that a browser extension is the way since it could too easily pull me back to the website. But I haven't really tried any solution like that for now, except for the simple keyword filters in RES.

For now, I'm using Cold Turkey to block some news subreddits, and News Feed Eradicator to remove all feeds. I can still visit specific subreddits and read threads. I don't think that's the right solution, but it helps me notice my patterns.

In any case, if you're interested in collaborating on this, let me know. It could be an interesting little project to work on.

I don't think the LLM idea is a great one, to be honest. For a while, I tried forums. ArsTechnica had a decent forum. Fewer people, better moderation. Even that had it's faults for me.

I'm still on the journey, but part of it was just... stopping with one of the communities that I was the worst on. Cold turkey. Habit Tracker app, don't comment there, did that for 90 days and then just stopped entirely.

I tried to block reddit recently. It's impossible. I wanted to find some nice headphones, everything leads back to Reddit recommendations. I still think I should delete my account, but I've had it for nearly 15 years.

But I can't see any good in replacing human interaction with LLMs. It sounds like a recipe for psychosis, in all honesty.