/tech/ - Technology

happy to see his hair is growing back

>>32460
neither is Saint Stallman in the files, but neither is Linus

>>32460
Saint stallman is just really autistic in semantics

>>32464
Not even esr?

>>32470
why would he be in the files, in the large scheme of things he's some random from UPenn, RMS would be more likely to appear since he's from MIT. I just don't think that Linux was big on Epstein's interests, seem like he was more focused on AI and NLP, thus Marvin Minsky and Chomsky were huge Epstein pals.

>>32439
When all else fails, make users keep a killfile! Vichan can hide by name, which is obviously of limited utility when anonymity is concerned, so maybe also implement basic glob or regex filters.

>>32455
>so maybe also implement basic glob or regex filters.
Filters are certainly what I was considering, but in a central censor to allow for more costly filters. Perhaps these could be togglable by the user to give a similar effect to the decentralized glob and regex filters. Obviously these posts wouldn't be federated, and would either be made hidden or removed.

The trouble is that I'm not sure the bots can do a good enough job. There are just so many ways to manipulate a conversation in a negative direction COINTELPRO - or even folks who do this work for free. I think I read that even detection of hate speech is something like a 70% accuracy affair for Facebook with much more developed tools at their disposal, and we're looking to detect far more subtle acts than hate speech.

I'm really at a loss for how to manage this, and it might be hindering my motivation to progress.

>>32461
Think there's really only two design objectives:
1. equal power to every post/poster
- no follower/following model.
- no ranking of posts by popularity.
- no names to discriminate content of users.
- (optional: following based on content via tags.)
2. quality content.
- access to a large collection of posts (federation - you could always make a new server but not you can get content on it).
- some restriction on the domain of acceptable discourse.

The fundamental contradiction is that without CIDR (centralization - and so no content) or usernames (reputation - and so unequal power) to distinguish posters I think there is no possibility of any sort of bans, and this makes creation of quality content difficult with today's level of technology.

>>32461
>a central censor to allow for more costly filters
Filters shouldn't be costly, unless there's something deeply wrong with your architecture (like emacs gnus scoring articles in a single thread by fetching them multiple times).
>The trouble is that I'm not sure the bots can do a good enough job.
They likely can't. This is why you give users the options to configure these filters, because it allows them to do aggressive filtering at their own risk. The more a filter or lack thereof bothers them personally, the more time they will spend on refining it, a textbook case of worse-is-better.

>>32471
>Filters shouldn't be costly.
Well, if we're using LLMs at all they would be from the user perspective.

>They likely can't. This is why you give users the options to configure these filters, because it allows them to do aggressive filtering at their own risk.
This tracks, it's a good idea. Having two tiers of filtering, one for basic federation (meeting basic social standards of the fediverse) and another tier with per instance filtering you could make it federate well, at least on the level of the Group. I also like the idea that sense we're working with trees of a single inReplyTo link per post you can go ahead and hide entire trees at once. Making bump ordering work with this is a tractable problem.

>>32450
yeah? it's somewhat expensive too because all the content they fetch is tokenized to infer the next n tokens, and they do it several times a minute with a cron job

>>32450
OpenClaw in general is a money hog.

>>32428
Hacked already: https://archive.md/PBKrW

>>32468
>detect and correct errors
It's plausible that these are error correction mechanisms. At the minimum there's trouble in everyone of these mechanisms in that the power to utilize them effectively is consolidated in the hands of the few. So naturally those people with power make their voice heard loudest, they run and fund their candidates, they get the best lawyers, even running all the studies (though one has more faith in science than many of these), and certainly apply all the scientific knowledge to their advantage.

In the imperial context this is even worse, if by some miracle, the people are able to develop and assert values opposed to power, chances are you just get economically strangled or overthrown, or both. And to paint a different picture as usual Singapore or China presents an alternative where there is a curtailment of freedom of speech, democratic elections, and judicial review, leaving rightfully only science. At the end of the day what counts is power, and competency.

it's been known to bundle adware for years, you need to compile cheat engine, it's not super hard, luckily

Infecting lazy script kiddies who can't even be fucked to compile their tools is based.

>>32422
No, you still don’t get it.

The entire rest of the fedi is NOT ‚connected and neutral‘, it’s just the extended HOA empire with slightly looser borders. The big ‚normal‘ instances (mas.to, troet.cafe, chaos.social, flick.social, etc.) are all run by the exact same German/NL/US progressive NGO-academic caste that runs mastodon.social. They all defederate the same targets on day one, they all have the same 40-point rules list, they all instantly gang-silence anyone who uses wrongthink keywords.

There is no ‚everyone else is connected‘ zone. There is only the Greater HOA Reich and then three or four isolated ghettos (chud, loli, and whatever’s left of the old weird fedi).

The whole promise of ‚join any instance, you’re still on the big network‘ is a lie. In practice there are exactly two networks that matter:

1. The progressive elite monoculture (90 % of active users, perfectly sealed, zero tolerance for dissent)
2. A handful of leper colonies where the outcasts were kindly allowed to go

Mastodon didn’t create a million free communities.
It created one gigantic, perfectly sanitized palace for the ruling class and then a couple of reservations for the undesirables so the palace dwellers can keep pretending they’re the tolerant ones.

Post too long. Click here to view the full text.

>>32425
Your premise hinges on the idea I implied that the HOA was just manstodon.social and not every too-large-to-moderate instance.

Any instance too large to reasonably moderate resorts to HOA tactics, as the trust that admins and federation as a mechanisn can actually protect against spam breaks down, causing hypervigilance to be an acceptable mindset.

The "The big ‚normal‘ instances" are like that because they hare hypermassive and can't stop expanding despite demonstrably being unable to handle the load.
>A handful of leper colonies where the outcasts
That's where the rest of fedi is. That's where all the activity is pubbing. As I keep telling you, everyone else either defederates from or limits these hypermassive instances due toetheir spam problems.
>were kindly allowed to go
You went into poteticism without thinking of what that represents. The HOA instances don't allowance anything to the rest of fedi, the rest of fedi are seperate websites that block or limit them. They are the twitterlerite lepur colony that keeps the rest of fedi free of that shit, not the other way around.

The reason why you see it the way you do is because you resonate with the culture of the HOA instances more, judging by how quickly you rejected picking a furry instance. You just want a twitter out of a technology that structurally evades being twitter on purpose, and breaks down when treated as a twitter.

>>32369
The real sin of mastodon is being a twitter clone. Twitter is a terrible format

>>32431
I like that misskey tries to break the mold a bit by replacing ⭐'s with emotes, but yeah frontends have only scratched the surface.

>>32369
Nice new phone

>>32357
wrong. they are my original ideas and i used llm to enhace them. learn the difference.

>>32355
fact: there is no such thing as "autonomy".

its pointless to discuss with idiots who don't know this.

>>32350
>We must stop fighting our nature and start engineering our environment.
Human nature is on the one hand a Spook (in the sense of being a relatively fixed restrictive idea), and on the other hand all that there is in human beings. There is no environment independent nature of man, yet all man does is a result of his biology, his nature, environment or even ideas can quickly constitute a second nature, but time is all that makes this second to a first nature. Human nature is something like the divine right of kings used to justify what currently exists which all sucks.

A question of mine, which relates, is why does the will exist? Or does it? Or does it have to? Why is there not simple wanting then doing? An LLM answered mostly with antiquated Spook filled notions of blame assignment in social and personal settings, and ranking of temporally separated wills, which seems like it could be completely automated. There was maybe something to do with planning, but I don't see the issue here either, you want so you plan and then you do.

>>32362
>Why does the will exist?
>Why is there not simple wanting then doing?
I suspect what we're talking about here is not the will but the idea of feelings of cognitive exertion even though there's no cognitive resource that is being exhausted. My guess is that our will needn't be effortful, this is even true for non-habitual choices, and happens when there is little internal conflict. So it turns out the vast majority of our human actions aren't effortful, but the question that follows is: why is this internal conflict not resolved effortlessly? I couldn't find a satisfactory answer to this, but it probably doesn't matter at all, the whole point of will is that you just ignore the strain. none-of-this-is-real.jpg

>Or does it have to?
Empirically people who think the will is a depletable resource tend to have a depletable will, while those who don't do not, so it's not out of the question that there could be some sort of psychological mechanism at play here. Not only a "spook" but a "spook" that artificially constrains us. My best guess presently however is that the main artificial constraint is not with the presence of the will but with the idea that we need to respond or even focus on strain. Real tough guy stuff: regular exercise, cold showers, meditation.

>>32350
I agree with >>32354 but this logic is just silly. Are all addictions myths? The fact is addiction exists, but the solution isn't to individually abstain from a powerful technology that is extremely integrated into daily life. We need to address the issues that make it addictive at the social (and for some individuals possibly medical) level, the same as other addictions. That will transform our relationship to these technologies. But saying it's not addictive is ridiculous when the big social/media companies literally run social experiments and engineer their platforms to be as addictive as possible and push the content that is specifically addictive to you as an individual.

>>32342
I think you're being overly cynical though? There's a lot you can reclaim out of your digital life as long as you have the interest to. Like, okay so basically I have a homserver at home, just an office pc I got for like £30, stuck a hardrive of 1TB alongside the preexisting SSD that came along with it. Stuck ubuntu linux into it with a simple ssh login with no display so that it sips as little power as possible.

I'm now able to stream MLP episodes form a nginx webserver with mpv I'm hosting in my own home with no reliance on Netflix. And I'm even storing some family photos. Just maybe as a bit of ecnouragement, here's my docker-compose file: https://codeberg.org/compuhsition/homeserver

>>32342
I'm also running a booru using shimmie2 alongside this too! Which I use to just host a few anime art references for fun. I make sure to block AI crawlers using anubis + I also use Cloudflare tunnel so it takes a lot of the bandwidth and security concerns away.

>>32342
All I'm saying is, as long you're willing to put in more effort, you can definitely be as uncensored as possible even in 2026. As long as it isn't like literally illegal.

>>32342
>because nobody will find my site!
Just link it on fedi or a minecraft server or whatever. No one will find you on anything if you don't share it with people you know. A website is just more stable and you can do more with a full HTML toolkit than what a social media post composer can provide. (rip cohost)

>>32189
did anyone else noticed, that substack is flooded with AI generated articles and substack doesn't seem to care? actually, it looks like they have completely embraced it.

>>32179
we have to retvrn to OStatus/GNU Social

>>32178
You can't reply to a redeet (or whatevr reddit post is called) from twitter, nor can you like a youtube video from facebook.
You can do several interactions with any misskey instance on any mastodon instance as long as they're federated, even being able to migrate between instances.

Like what more do you think it was promising to be? The most decentralized option is just making a website, and every individual website is centralized.

>>32179
Oh your issue is right wingers can't spam peter-theilisms, gore and CSAM to every existing instance, and that it isn't just recreating the sophism machine.

>>32365
Solved by posting in all of them

>>32353
>Could I detect those nasty AI cameras used by law enforcement
Maybe you can go out now and make experiments, put the device you are going to use promiscious mode and check if they have some fingerprint you could use. Then you would have to deploy a considerable amount of them (how are you going to finance it?) which I find it hard. Are you in an org? You would help more that way than with individual action.

>>32360
Also, have you tried considering asking on lainchan? If you do try to make a good OP post so it doesn't get flagged as soykaf, lurk there a bit or ask on their IRC (I guess)