/tech/ - Technology

>>32143
I'll try it out.

>>32140
Meh, I was excited about Lemmy at first but it's just hugboxes all over again. Honestly, I think upvotes/retweets/likes just ruin the whole thing from the getgo, it just breeds the insufferable "hot take" environment

Fully Parasocially Detached with Maximum Algo Exploit
A radical and extremely rare way of using X (as of January 2026)There exists a tiny minority of X users who simultaneously and uncompromisingly pursue two goals:Letting the algorithm work for them as intensively and honestly as possible –
while at the same time radically erasing every trace of parasocial connection, no matter how faint.The core of this approach can be summed up in one single sentence:
“I want the feed to understand me better than any human ever could – but I want to be practically non-existent to every human being on this platform.”The technical implementation is relatively straightforward, yet extremely consistent down to the smallest detail:
The account remains permanently protected. There are zero followed accounts – forever. Anyone who attempts to follow is instantly blocked. Likes are set to private, there are no replies, no quotes, no reposts, no mentions. The “For you” tab is the only one ever opened. And one’s own behavior toward content is kept as honest and impulsive as possible – no counter-steering, no strategic clicking.The result is paradoxical: After a few months, the personalization quality often surpasses what most normal power users with hundreds of follows ever achieve. At the same time, social visibility and parasocial attachment are almost completely eliminated.Who actually does this over a long period (not just as a short experiment)?Most often, these are people with a very strong need for protection from closeness. Frequently you find avoidant attachment styles combined with an extremely high need for autonomy and sovereignty. Some come from long, intense phases in toxic online communities, fandoms, or political camps and simply never want to be caught again in any form of expectation or group dynamic. Others are principled privacy maximalists – often with libertarian, crypto-adjacent, or radically data-protection-oriented backgrounds – who fundamentally reject follows and followers as an undeserved, structural form of attention economy.After many months, the few who really stick with it tend to say similar things:
The feed has become uncannily accurate, sometimes uncomfortably accurate.
It feels like a second, completely honest consciousness that belongs only to oneself.
The first months were chaotic and poor, but from around the six-month mark it got really interesting.
One has discovered things about oneself that one didn’t want Post too long. Click here to view the full text.

The Digital Scavenger Hunt: Why I’m Abandoning the "For You" Feed

The modern internet is designed to be convenient. Algorithms on platforms like X tirelessly curate "For You" feeds, aiming to show us exactly what we want to see. But this convenience comes at a price: we are trapped in a feedback loop of the familiar, governed by engagement metrics rather than genuine discovery.
To find something truly unique, you have to break the machine. I’ve started using a different method: bypassing the algorithmic gatekeepers entirely by manual searching via xcancel/Nitter.

X is essentially a massive, chaotic landfill of data, a mountain of digital trash. The platform's algorithm is designed to surface what is popular or provocative, but it is fundamentally incapable of recognizing a "hidden gem." It prioritizes the loud over the profound.
When you use xcancel to search for specific, niche (random?) keywords without being logged into a personalized profile, the experience is jarring. The vast majority of what you see is irrelevant noise. However, this is where the magic happens. By manually sifting through the debris, you encounter "pearls", eccentric thoughts, obscure primary sources, or niche conversations- that an algorithm would have filtered out for being too "irrelevant" to your profile.
This is the fundamental difference between randomized discovery and algorithmic personalization. One gives you a mirror of your own interests; the other gives you a shovel and a map to the unknown. It is tedious, manual work, but in a world of polished, predictable feeds, it is the only way to find the treasures buried deep within the data pile.

X algo now open

>>32276
>>32275
now it is official, why you should avoid the algo

>>32219
A trait everyone who has AI psychosis shares is that the delusion that LLMs actually experience shit, this is no different from the magical unlock prompts, LLMs really start spouting total gibberish when you give them greek symbols, anyway here's qwen coder (i was too lazy to download another model) solving the very real muller/friedland hypothesis after feeding it the "operating principles"

>>32227
very cool thanks qwen

i dunno if this is the right place to talk about it but it's starting to bother me a little bit with this rampocalypse shit that no one seems to put together that the reason why supply chains are so fragile is due to US export controls on chinese semiconductor companies lol, it's all AI this AI that and I hate Sam Altman, and I get it, Sam Altman is a pedophile sister diddler, but it's also like you can clearly see who is molesting the supply chains, right?

Teste

>>32187
Over here the anti-AI sentiment comes from Historical Materialism. Over there the anti-AI sentiment is just 'change bad' though.

What are they? Can you provide any example? Are they like electron apps but for android?

>>32253
I've only seen them mentioned in the context of websites, because they're the kind of sites where back and forward navigation don't work. It means every function on the site itself is implemented as part of a webapp instead of a link, basically the same type of code that would be put into an electron app running in your browser.

>>32255
Oh I fucking hate not being able to go backwards/forwards, they are usually super laggy and shitty
Nothing should be really done on the phone besides texting, photos, calls and notes (and maybe payments)
Hard pass

>>32271
lol what?

>>32256
>I kept getting bogged down before in my dislike of in effect having two different account systems: one username, and password pair for moderation (session based - in part for protected views), and a (optional) trip, and password pair (form based) for posting.
I guess the merger we've just been discussion is roughly that there could be session based accounts for Actors (how you get a trip) and deletion passwords for posts, with the password hash. You can delete or edit either by the post password hash or by the session login.

>>32256
>you could use name for the display name with preferredUsername set to the secure trip which is "unique"
In the general case there are normalfags on the site, which wouldn't want to link to user pages by a tripcode.
>An advantage of the tags over board approach is that it makes clear which Group is posting, the only one that exists locally, if you want to reach out into the rest of the fediverse for posting.
In my planned approach boards are immutable, supersedable anchors, that allow threads to link to them. Therefore a different instance could create a post linking to a board or a thread (anything allowing replies) and every node in sync will render it on the board in accordance with the implicit graph structure.

In this case unstructured, tagged posts should probably only be propagated on small nodes or use a local tag, which may be shown as a "timeline".

>>32258
>for normalcomards
Well, you're right this isn't ideal. Mentions would be uglier even if you had autocomplete from follows/thread actors and info onmouseover.

>immutable, supersedable anchors, that allow threads to link to them
Excellent. But is it simpler, and more interoperable than just having a custom of mutual following for Group actors?

>>32259
>custom of mutual following
I don't think follows should be part of the protocol. My scheme would use replication structure i.e. readers are either downstream nodes or exposed out-of-band by another node. Prose is highly redundant, so i think native lz4 compression would prevent even a several decade-old network run on top of the protocol from reaching the same multi-terabyte data sizes as usenet. If storage size was to blow up though, node admins should still be able to pivot to a more frugal replication policy, like most activitypub instances have by design.

>>32261
>I don't think follows should be part of the protocol. My scheme would use replication structure i.e. readers are either downstream nodes or exposed out-of-band by another node.
I don't fully understand, and admittedly my abilities to implement complex programs are limited, if this would in fact be one.

>native lz4 compression
This is an important point. I believe you can do this for postgres rows.

I do now have a program (technically called "arsvia-redux") which has the creation of Person objects for login and signup using HTMX, sessions, and CSRFTokens. It needs testing, review, and revision. I'm not sure I really have the energy to dedicate to it or not.

Fedilab is pretty good. It doesn't do twitter frontends, but you find yourself a good activitypub instance and it'll do infinite scroll.

>>32265
Simply cease to desire to use twitter then.

alright, i found a solution 👇

https://xcancel.com/elonmusk,SarcasmSilly,yanisvaroufakis,glitch_prod,eliteobsolete,Kalshi,sama,juddrosenblatt,oishikaiju,leftypol_org,curtis_yarvin,neet_sol,iluminatibot,OldInternetFeel,beff_jezos,evacide,naomibrockwell,9GAG

>>32265
rude! no bullying allowed!

>>32240
Maybe I'm just really good at shapes or because I'm not good at touch typing yet, but I find it much easier to just drag something with hard stops and click next, than to drag something gradiential, question if it's aligned right, guess what the characters are through a bullshit font, and type it in.

>>32244
please, just stop justifying these horrible captchas. the only reason they exist, is to sell the stupid pass. we are witnessing the enshitification of imageboards in real time and anyone who mentions it, gets banned.

The captchas that haven't already died from the rather draconian third world captcha farms paying pennies per captcha are destined to be destroyed by LLMs. The only future is in filtering the content before it can even be posted with LLMs.

>>32239
anyone bought these? genuinely curious

>tenbux hiroshimoot edition
Rest in piss.

There being only a group frees the individual from responsibility and accountability, while the simple algorithms for sorting give near equal power to every commentator. I don't like the ephemerality of posts, seems kind of pointless, and the interface makes it difficult to collect and repost in later discussion. Resolving this would make a decent enough hypermedia system for regular power users. It's also be cool if the deletion password could be used to draft new versions of old posts, but this is more complex.

I guess feeds might have lead to the development of the sort of political culture we have today. A better chunk of the the actual "facts" come from the same places as always - there's some control here. But there's no real mechanism to hoist the right view on to the people. Does the undercurrent of the American Dream (false-consciousness) and Manifest Destiny (settler colonialism), surface in all its detail as the political reality? The wasps build their home together, but without there being any real control.

Of course this is old hat. Ideally there would be some mechanism for something like left-wing control as a replacement for feeds. Something yet still with outreach potential… Any ideas?

>>32230
>Ideally there would be some mechanism for something like left-wing control
The only think I can think of is to run an LLM censor on all the posts while otherwise operating in way I described imageboards as operating before: there's only the group and no individuals, and equal power is given to every commentator.

I'm thinking about how imageboards like leftypol could still have a reason to exist in the future, especially in the context that AI has made marxist interpretations of reality and discussions largely obsolete. Since AI draws its knowledge from the internet, including imageboards, leftypol would need to provide certain original content that AI cannot synthesize. but what kind of content would that be?

currently, all discussions on leftypol are very easily reproducable through AI. there are actually very few patterns on this site, which constantly are being regurgitated. what is it, what humans on leftypol can offer, what AI can't? this the question which will decide the fate of this site.

>>32245
we can meetup and suck each other's dicks. think about it

>>32245
Well there are still unknown unknowns, consensus building/following, perspectives that don't show through a computer generated average, and probably a host of other things.

seems like they plan to leverage some built-in AI moderation thing which may or may not exist to woo investors

>>32213
Had an idea to use this for my personal project, main trouble is that even once you make the filters are little more liberal you've got something like a 70% catch rate for harmful content.

>new
Not sure if shitpost. Digg launched in 2004 originally, but I didn't know about the relaunch, so thanks anyway.

Did you do anything to get banned or was it a range thing?

>>32205

Range I think, first time going to the site and was already blocked from making an account.

>>32204
>calls brainfuck brain****
why are they like this?

Based on the abstract of the article that you have posted, that study does not establish causality. It does not indicate that "short-form video has a significant negative impact on attentional functions and self-control." It says there is a correlation between the tendency to get addicted to short videos and weaker attentional function and self-control. That is, if you lack self control and have bad attention, you are more likely to get addicted to short videos.

>>32192
You're correct, and I'm having difficulty finding any quality studies which truly demonstrate this. The discussion seems to relate back to a single theoretical source: "Amusing Ourselves to Death" by Postman. We know further that far transfer does not exist, even if mindfulness and WM training work has had significant positive impacts on attentional tasks, and WM tasks respectively (near transfer). Further anecdotally I know people who use social media and read more books than me. Maybe the only change for good or bad in switching media forms is the sort of information you gain for the time invested, and how this integrates into the cognitive map.

>>32194
It's a moral panic. This chart from the substack article is said to be explained by the introduction of smartphones and there's no mention of austerity… Laughable. Obviously it's the masses that need to change their moral character and not the politicians who have been looting our public institutions for decades.

>>32195
>It's a moral panic.
>austerity
This is convincing thank you. Just to give some concretes attached is the decline in education spending as a percentage of GDP which peaked in Europe before WW1 decreasing around 50% and in the US just after WW2 decreasing around 25% (I don't mean to distract by anchoring with the wars but it might).

This is also ignoring the fact that at least in the US, and perhaps in European countries too (I'm thinking of private schooling in particular) there is systematic discrimination against poor children, and young adults in funding of schools and education. That being said it may be that decreases in funding have primarily been depriving the elite, this seems highly unlikely however.