returncatalogbottom
rules1. you must be 18+ to use this site 2. no NSFW/gore 3. no bigotry or slurs 4. if staff don't like your post they may delete it or ban you

anon
genuinely dont see how people can find that stuff appealing, it's just too far into the realm of absurdity
anon
>>105359
is there anything to hosting a website? i zoned out during it courses and am at the point where computers are magic boxes
>>105370>>105368
Anonymous
>>105364
hyperherm kitty thinks thats called on those nsfw sites
anon
>>105366
computers are magic boxes you just have to fins the programs written by severe autists instead of corporate normoids
anon
They are the real magicians
anon
uh shit its the fuzz, cheese it gang
anon
#uwu (USER WAS UWU'D FOR THIS POST)
anon## admin
theirs is as far as i know the first ever liveposting software so i kind of just had to wing it
anon
>>105370
>rust
the language syntax looks weird
anon## admin
their master branch is a failed 6-year-long rust rewrite attempt that doesn't really work
anon## admin
so we're running the original version which is pure Go
anon## admin
and TypeScript for the frontend
fujo
how did they not manage to push out a rewrite in 6 years..
anon## admin
full rewrites often go like that
anon## admin
yes
fujo
i mean, a lot of software also just turns out worse after a rewrite, gcc for example.
anon## admin
i am not sure if meguca/shamichan v7 (the rust version name) would've turned out better but it was just a doomed effort
>>105385
anon
>>105384
Somewhat related? With the public release of DeepSeek, do you have any intention of letting users choose what AI model to use with the #ai command?
- Would a submarine make a great spaceship?
- How can I get users to RTFM?

more hot questions
question feed
command?
>>105390>>105394
SweetMelancholy
oh my NEET bones ache
SweetMelancholy
from moving and walking ahh my hips hurt ughhhh tiredd
SweetMelancholy
dont sit down for 6 months it hurts to get up
i like sad things
>>105385
i can answer that question having been on a submarine
fujo
>>105381
i think the whole article boils down to "code re-use is hard"
which is an interesting statement, i don't think code re-use is hard per se, it just is applied very wrong;
the more fundamental a concept is, the easier it gets to re-use code, you can see this with like, conversion of data between 2 formats that are easy to grasp, such code is widely re-used.
>>105393
i like sad things
pro: air tight con: water cooled so you still just accumulate heat until the pressure kills you. vaccum is the best insulation possible, no environment to exchange heat with
anon## admin
>>105385
i don't like spoiling new features but: all AI commands have been replaced with deepseek as of yesterday. it'll be deployed here within a week probably. there are just uhhh reasons why i work on changes for months before deploying any of them
>>105396>>105400
i like sad things
(jk)
anon
#ai common trans girl names

1. Ashley
2. Taylor
3. Madison
4. Samantha
5. Hailey/Haley
6. Alexis
7. Chloe
8. Riley
9. Morgan
10 . Sydney/Sidney
common trans girl names
anon## admin
>>105396
it was added in dev weeks ago so it's still part of the same bundle of changes waiting to be deployed
Anonymous
god i fucking hate having an adams apple, i swear i feel constant discomfort from it being there, even physically
anon## admin
i mentioned it before but catbounce required tons of code to make it work optimally. scroll-lock made the cat bounce less effectively
>>105404
i like sad things
you want a tip?? link ko fi :)
>>105403
anon
>>105401
web dev was a mistake - if you want to talk to other people use e-mail or irc
i like sad things
>>105403
yea look up ko fi its like the most open source least evil tip jar all the indy game devs and mod creators use it
anon## admin
i've seen it but i wasn't sure if the recipient's real name might show up in a thing or something
anon## admin
i'll set one up. for now we just accept cryptocurrency
i like sad things
unfortunately I'm a cryptolet
anon## admin
>>105393
this but unironically
i like sad things
i have no wallet, I'm not on the block chain, zero shitcoins, not sigma grindset, no doge, graphics card just idling when it could be mining
>>105412
i like sad things
yea looking it up ko fi lets you do essentially a branded storefront no need to put ur name there you cab just say brainworm.dev and you can even set up patreon style recurring donations or a little web store
>>105418
anon
>>105410
bitcoin mining on consumer hardware seems like an easy way to ruin new stuff quickly
>>105414
i like sad things
well actually no, graphics cards are insanely good for mining its just like... if you have one why burn it out for mining even you own it to game lol.
i like sad things
all the biggest crypto farms use nvidia cards
>>105419
fujo
>>105413
this has been my experience trying to use ai for anything useful, like all the time.
idk if im too dumb to write good prompts, or if my questions need "real thought" that ai just doesn't do
>>105706
!nIOh.loris
>>105416
no wonder the gfx cards are so expensive
anon## admin
>>105413
4o-mini isn't great. try claude 3.5 sonnet or deepseek r1
>>105422
anon## admin
or maybe they're all bad at julia
fujo
>>105420
claude's code doesn't even parse, it gives me syntax errors
>>105423
anon## admin
>>105422
damn. it's generally considered the best
anon## admin
have you tried it for more popular languages
anon## admin
for me, claude regularly writes 300+ lines of code that are correct on the first generation
fujo
yea, i tried it for generating some c++, but everytime i do that, i look at the output and im just "meh, this doesnt seem that good" and rewrite it myself anyway
>>105427
anon## admin
>>105426
on twitter i saw people saying all the models are bad at C++
anon## admin
i think they excel at python, javascript/typescript, and go
fujo
can't speak for any of that, i don't do anything with those languages
anon## admin
i guess give r1 a try just in case
fujo
oh also, do you do the inference for r1 locally, or do you pay for their api
anon## admin
it would be extremely economically bad for me to run any of them locally. even if we had a million daily active users probably
>>105446
anon## admin
libpol v2 uses deepseek v3 for all text AI and for images defaults to stable diffusion turbo. and is SD lightning. is fluz schnell, is flux dev, is rewrite v3, and then we have like 12 more for some reason
>>105435>>105438
anon## admin
i just got rid of GPT completely. i like deepseek's responses more, and it's way cheaper
>>105437
anon## admin
>>105435
libpol v2 is still in dev but will be deployed soon. it contains many other changes
!nIOh.loris
>>105434
>way cheaper
LMAO
western ai development has been btfo
anon
>>105433
absolutely based we love deepseek we love china and we FUCKING LOVE THE FREE MARKET
>>105439
anon## admin
>>105438
yeah i'm happy to use whatever is best and deepseek seems like a great company. i don't think they're like a CCP operation or whatever
>>105440>>105442
!nIOh.loris
>>105441
#this (this)
also, the fact that they made it fucking OPEN SOURCE is so fuking good. many nations of the world can develop their own ai stuff and also any person with a good gfx card can do their own stuff too
fujo
i think ai is deeply worrying
>>105447
anon## admin
>>105442
well this site is called libpol.org and the CCP is anti-liberal democracy
anon
>>105432
doesn't deepseek still require hundreds of GB of VRAM for a local copy?
>>105448
anon## admin
>>105446
yeah i just mean their API is very cheap. locally i think it's still cheaper than some things but the API makes the most sense for us
anon
>>105447
i think its bad for ur brainn
anon
i think privately funded ai is deeply concerning, i think open source ai is fucking based and epico
>>105451
anon## admin
>>105450
the main argument against it beyond like national security stuff is the existential risk argument. but either way that's not gonna apply to these lines of models
>>105469
anon## admin
like libpol is closed source mainly for security/arguably existential reasons, so i can understand that argument
>>105453>>105454
anon
>>105452
yeah no dont worry ive heard your decisions for that and i think you're completly making the right choice. It's more that like basic infrastructure type things should be open source. You should be able to own a bakery and keep it private but the road to get to the bakery should belong to the people
>>105455
fujo
>>105452
existential reasons in terms of like "people shouldn't be able to host a knock off version"?
or "security" in a "security thru obscurity" way
>>105456
anon## admin
>>105453
so similar arguments for locking down AI can, in theory, make sense
anon## admin
>>105454
the latter sort of though not exactly. well for AI it's complicated i guess
fujo
>>105447
but yea, i mostly mean something like this:
search engines get worse and worse, so people tend to gravitate towards using AI instead of search engines.
historically, there have been countless of debates that were about "search engine neutrality", so google censoring stuff would've stirred up quite a large debate.
we're not having the same debate for AI models, and if we do, it's mostly coming from a weird rightoid standpoint.
anon
>>105460
its kindof pizza dough but mostly i just eat it with salt and olive oil and olives on top
>>105466
anon
its based on adam raguseas recipe on youtube but i made alot of changes
anon
its really high hydration (even tho it doesnt look like it atm) and it sits in the fridge to ferment for like a week ish
anon
but i end up eating it before the weeks done mostly xd
anon
>>105451
I kind of have a hard time imagining the existential risk part, i think a disinterested actor created by humans has basically no motivation to tear down humanity
>>105471>>105472
anon
If it's really really smart which is the whole assumption, and it has some kind of goal that doesn't particularly align with human interests, I don't see how le enslavement of humanity would ever be productive to that end. Seems like the better option would almost always be either completely leave humans alone, or make their society really prosperous but force them to contribute some kind of creative/developmental labor as taxes
anon## admin
>>105469
i am ambivalent/agnostic on it. yudkowsky's perspective is it will have no active desire to end humanity, it'll just happen as a byproduct of other actions and won't care
anon
>>105469
i dont think this is what people mean by existential risk, but i think that it will automate away everything that makes us human and then we are just not human or whatever
>>105473
anon
>>105472
I mean if it's in charge of everything it seems weird for it to have some kind of intrinsic desire to like, create art for humans
anon
to the point of replacing all human labor in that area
anon
And the physical labor stuff is just silly, there's no job that a human can do better than a robot and the earth isnt a very good source of any raw material compared to other bodies in space
>>105481
anon
i think friction is what causes you to be a self reflecting conscious agent
>>105478
anon
So cooersive labor just seems silly to me, idk
anon
>>105476
hm i dont really understand sorry, can you explain?
anon
Like i just mean if the ai decides the goal of its existance is to get bigger and make itself more chips and get smarter, the best thing for it to do is fuck off to asteroids with extremely high grade ore. earth is not a good source of any of those things, and humans are not good at doing that labor
>>105482
anon## admin
>>105475
yeah i don't think the existential risk people are worried about mass enslavement. i think a typical scenario is like "AI decides putting some thing in the atmosphere is the best way to achieve some goal and that also causes a billion people to die"
>>105483
anon
>>105479
no i don't really think about it like that, i don't think that the ai has any intrinsic motivations, i'm not interested in that sort of existential thing. instead i think like if it gets really good at giving you whatever you want with no friction, then i think you just turn to a life of infinite escapism or whatever, and i don't think that that is exactly a conscious life because there's increasingly little impetus to reflect on things
>>105489
anon
>>105481
I guess i understand but at the same time i can't think of any end where that's particularly useful to do on EARTH instead of somewhere else
>>105485
anon## admin
>>105483
yeah that was kind of a contrived scenario by me. i don't really know the actual odds. just if some thing is insanely smart and resourceful and powerful and theoretically might not have deep empathy for human life/moral worth, bad things are possible even if it's not intending to harm humans
>>105487
anon
And like idk, humans are very unique and took a gazillion years to come into existance, if there is even some infitessimally useful thing they can provide like creativity or extreme long term robust data storage in dna or something, the best way to extract that from them is just to make their society prosperous
anon
>>105485
I guess, but i still think a disintrested agent keeps humans around and doing well on the off-chance they do become useful in future, because it's really really hard to put the toothpaste back in the tube on that one.
>>105488
anon## admin
>>105487
for a while it makes sense but if you extend to like 50k years it may not really see any benefit to humans existing. or dumber AIs existing. most humans don't care that much about chimpanzees and dolphins and their societies
>>105490
anon
>>105482
like if you think about art as this thing that makes us human, that's the trope, i don't think that in art, having more power to create exactly what you want perfectly accurately is a thing that inspires artistic depth. instead i think that art comes from the limitations of the medium. just taking photographs simply is much easier at capturing how someone looks than painting them, but i think just taking a photograph of them creates something that has a lot less depth, unless you're pushing the boundaries of the medium, compared to painting. so that's sort of how i understand friction
anon
>>105488
But in that case i only see it killing us if we're in the way of some resource it wants and there's no resources that are easier to find here than elsewhere on a mass scale
>>105492
anon
Like yeah if you want only 100 kilotons of gold to make chips with, sure, stealing all of humanity's refined goal is an efficient way to do that. but if you want 100,000,000 kilotons it just stops being a reasonable thing to do
anon## admin
>>105490
yeah it seems fairly likely that for a long time it would basically just ignore humans but there are so many possibilities in the distant future that it's hard to rule out it discovering that like some compound is super valuable and it adds it to or removes it from things or really just any large-scale engineering action we can't conceive of
>>105493>>105495
anon
>>105492
I mean fair enough but I think at that point its harvesting the offshoots of recently popped supernovae where everything "rare" is common, or exoplanets around stars that did that a few hundred million years ago so it coalesces a bit and makes your job easier
anon## admin
like how big construction projects can harm ecosystems without us realizing it or really caring
anon
>>105492
and i just think that if it has unlimited foresight which seems like a reasonable thing to asume about a quasi omniscient agent, then it would want to keep humans around, because if u realize 100 million years down the line that actually having a working human society is at all beneficial, ur out of luck, its really fucking hard to remake them
>>105496>>105498
anon## admin
>>105495
it may know it will likely or definitely harm humans but it might be missing some other part that makes it choose a different course of action
anon
the real question is will friendship is optimal come true
anon## admin
>>105495
i think it probably wouldn't be beneficial, for a sufficiently advanced entity
>>105500
anon## admin
or only as a spectacle or out of empathy, like humans rescuing endangered species
anon
>>105498
i mean idk why not? im just saying we, ourselves, are the rarest resource of all, and u dont just destroy a resource as a disintrested amoral agent
>>105502
anon
What if it wants to do like biotics research because they are way more compact than chips, easier to keep your 2.5 billion year head start around than start from scratch
anon## admin
>>105500
what about a world with 1 trillion superhuman entities and 20 billion humans. they'd consider other superintelligences a more useful resource to conserve and work with
>>105503
anon
>>105502
i know but we are pretty dumb in this relative scenario and even as dumb dumbs we collectively try really hard to NOT wipe out random sahara ant species #292947, not with even any notion that theyre useful, just because, you cant remake them if they all die. I think thats a very reasonable ethic for an amoral agent to have. dont take the toothpaste out of the tube because its not going back in
>>105504
anon## admin
>>105503
yes but i think we can't assume by default all new intelligences will share that logic. i think...
>>105508
na-bee-eh (nadie)
I'm sad to announce that Glegle has passed away last night
>>105506>>105514
anon## admin
#ohno
anon
>>105504
idk i think resource preservation would be an essential quality of intelligence
>>105511
anon
And i think humans are just a very difficult to replace resource
anon
iunnno
anon## admin
>>105508
i think after a certain point it may not consider humans that interesting to preserve. but it might
>>105513
anon## admin
also it could just keep a few around in a lab and only preserve those for research
anon
>>105511
idk i think anything which takes conditions so perfect they seem to only arise once within the observable universe and take several billion years to develop are hard to justify getting rid of, even if it's a "just in case" type thing.
>>105515
anon## admin
>>105513
99.9% of agents could believe that. it would only take one that just doesn't for some reason (though i don't know how likely it is there'll be tons of godlike beings all around)
>>105518
!nIOh.loris
i like the gnr version
>>105522
anon
>>105515
i mean sure yeah i just think it's unlikely that life could ever be seen as unuseful
anon
Some things it would 100% want to copy like dna, its reaching close to atomic limits for data storage/stability
anon
Which would be useful for any self improving agent
anon
maybe u dont need to keep the meat suits around it sure but thats part of the appeal of DNA its information u can store and come back a few thousand years later and instead of degrading you get a shit ton of new copies of it
>>105527
anon
>>105517
axl rose doesn't seem like he has the voice for this
anon
Even if something locally disastrous happens to some portion of it
anon
i don't think anything is guaranteed with any speculation of the future like this but i do think on average a superhuman intelligence who takes over humanity would probably be altruistic. i would take the odds any day, i think it's more likely to want to help humans than the average human leader certainly
anon## admin
>>105521
i don't think there'll be good technological/science reasons to preserve humans for that long. they'll surely invent a better version of DNA. i think in the beginning period they'll have good reason to keep humans around, but in terms of total utility they'll extract what they need and then the only real reason will be emotional/compassion/respect reasons
>>105528
anon
>>105527
Doubt tbh, earth is a planet sized computer thats already been crunching away at the data compactness and reproducibility problem for 2 billion years, there's no way its not useful. even if you can make better epedrine you don't stop studying the ephedra plant because the primitive baseline you started with still has insights into how to improve it
>>105530>>105533>>105529
anon## admin
>>105528
DNA and proteins are very impressive but this scenario basically assumes that some AIs have cracked the code and can recursively self-improve/engineer/research far better and more efficient than evolution and beat evolution in all categories
>>105532>>105531
anon## admin
>>105529
i don't think this scenario will happen within the next 25 years but i think it could within the next 250 years
anon
>>105529
i mean just fundementally our way to create these things *now* is basically evolution
anon
>>105528
they can retain samples of DNA, and maybe retain a few living humans in a lab. i don't think it'll be that important to them that human society persists unless they have a moral feeling about it
>>105534
anon
>>105533
But that takes effort and shit, why not just leave them alone and they preserve themselves
>>105536
anon
If you have a giant big vat of feedstock, and you have an infinitely large warehouse, and you think you ever want a drop of feedstock in the future, why do you destroy that vat and instead just keep a little vial around? why not keep the vat?
anon## admin
>>105534
they might think the costs outweigh the benefits (if indeed there really are benefits, which i'm not certain of). they might be able to build things much more effectively if they don't need to worry about externalities to humans
>>105537
anon## admin
>>105537
i dunno. they unintentionally kill a few humans, human societies rise up and shut them down/disempower them. then it'd be better to just get rid of most of them all at once (though this is shifting the scenario to one of intentional extermination)
>>105540
anon
>>105538
i think this is the one reason theyd want to attack us, if we posed some kind of threat to THEM. but even in that case i still think the better and smarter path is to give them what they want with super minimal effort and then they dont even want to fight you in the first place
>>105543
anon## admin
>>105540
you'd need to talk to/read yudkowsky (warning: i do not recommend) to get some of his ideas about why he thinks the incidental existential risk scenario is the most probable
>>105544
anon
>>105543
>read yudkowsky
i'm not reading harry potter fanfiction
anon
if le big brother 1984 robit is like "actually i love and care about you and want your lives to be happy and free of worries" why would u ever attack it :P
anon
>>105542
>american football plays while the player fails to ask out another gle
anon
Easier to prevent the fight before it happens by making everyone happy
anon
The things we need to be happy, on the scale of like grand scheme resources in the universe, or even just in our solar system, are fucking tiny and negligible like u dont even need to put any resources in u just need a tiny fraction of ur intelligence to tell us how to organize stuff and efficiently self sustain
>>105551
anon
i think hunans assume antagonism out of an incomprehensible entity precisely *because* of our biological drives which that entity wouldnt have
anon## admin
>>105548
the same is true of many animal species but many go extinct due to our action or inaction
>>105554
fujo
sleep is no replacement for caffeine
anon
>>105551
agreed but i think its the dumb humans that make shit go extinct and the smart ones who fight to prevent it so i assume that trend continues as intelligence increases
>>105556
anon## admin
>>105554
i think there is a correlation but i think it's mainly just "cares" vs. "does not care" and that doesn't always map to intelligent/unintelligent
>>105559>>105561
anon
>>105556
i think if u look at the smartest people who have ever existed, the like prodigy freaks who do a generation of work on their own, they almost universally are anti-war, pro conservation, etc
>>105562
anon
>>105556
this desu morality does not equal intelligence
anon## admin
>>105559
at the extremes i think the correlation is stronger yeah. but there are some very intelligent people who are like far-right and sociopathic sometimes, even if the vast majority of very intelligent people are the opposite
>>105570
anon
its hard for me to imagine that trend doesnt continue. humans are a really rare resource and damaging any resource irreparribly even one you dont think you'll need any time soon/ever is not a smart idea i dont think
shrimp
shocking!
anon
>>105562
i mean yeah sure but i think those cases can almost always be ascribed to biological urges that are overactive in that person like fear of unknown that i dont think would be present in something without an inbuilt fear of death
>>105572
anon## admin
>>105570
i think most ideology is probably mostly a neurological feature in most people tbh. but then you can imagine an artificial neurology that for some reason is very smart in most things but has things we'd consider dumb or irrational that lead to harm
>>105574
shrimp
oh right
anon
>>105572
but i don't think loving compassion is an ideology i think it's a rational and amoral decision
>>105576
anon
Its always more useful for everyone to not waste resources hurting ppl
anon## admin
>>105574
i don't know, like if all dolphins went extinct right now, the main drive that would want to prevent it is due to compassion, not "dolphins are useful for us and progress/advancement"

>>105577
they are but like not so useful that a lot of people are going to deeply care for that reason
>>105577
shrimp
what have dolphins ever done for me? nothing
anon
but i think part of like environmental compassion specifically is "maybe dolphins will be SUPER useful in future, we have no way of knowning, and if we get rid of all of them, that's it and we can't undo that. so just in case let's preserve the potential resource"
>>105580
anon## admin
>>105579
well there's like the empirical question of the second-order effects to the ecosystem if dolphins disappear but then like take a scenario where maybe we're certain the second-order/nth-order effects won't be that net negative for humans
>>105581
anon
>>105580
yeah even in that case i think the purely rational, non-ethical correct decision is still to preserve dolphins
>>105582
anon## admin
>>105581
by default it's better to preserve than not preserve, but if it's just like some companies doing some things that slowly kill them and not enough impactful people care enough to slow that down and they eventually die out....that sort of thing is pretty plausible and has happened and will happen a lot, and the main thing against it is people having a sort of compassion that i think is not necessarily due to rationality or intelligence or prediction of future benefits/outcomes/chain reactions
>>105583
anon
>>105582
exactly, always better to preserve, and i think there's no scenario where some drive that the existance of humanity truly gets in the way of where its a worthwhile trade to get rid of them
anon
If iron was really rare in the universe and getting at the planets corr was useful then i think theres more arguments that it could be worth the tradeoff but literally nothing on earth is useful like that, all of it is not only replaceable but actually eclipsed by more efficient sources elsewhere in the universe
anon## admin
i probably shouldn't be the one trying to argue this. i wish i could get some like AI researchers and writers on this site.....livechat AI discussions/debates would be fun
>>105591
anon## admin
i'm generally pro-AI and not really in the AI doomer camp. i just think like "oh, 1% chance that, conditional on there being a hyperintelligent entity that's orders of magnitude smarter than the smartest human and it also is able to be autonomous and have tons of money and resources and power, it might lead to deaths of tons of humans? that's believable"
>>105587>>105588
fujo
>>105586
i always feel like im walking a super thin line, being like "pro ai" in a technology sense, but not in the sense of "we should throw it on everything we can!!"
>>105589
anon
>>105586
yeah kinda funny lmao, im reasonably more anti-ai than the average person, our positions on the extreme outcomes of this scenario are kind of flipped from our opinions on the current pragmatic end of the debate
>>105593
anon## admin
in terms of a practical sense i think the AI slopification of things is often very annoying and useless. i like it more in the abstract. and i do personally find it useful for programming (even if you don't; sry)
anon
>>105585
are there any actual ai experts who talk about this in media? all the talking heads I see are nocoders
>>105592
anon## admin
>>105591
i don't know, i don't think i pay attention to media really
>>105595
anon## admin
>>105588
that's what makes debates fun sometimes. or like this type of debate vs. the teamism stuff where no one would adopt a position that doesn't vibe right with their camp
anon
btw this site has the best debates ive ever had on any site theres nowhere else i know of where i could have this debate and not have it result in anger and shit flinging, appreciate u and ur work opwormulon <3
>>105596
anon## admin
>>105594
well i guess there are a few factors:

- part of my theory/bet about liveposting is that liveposting can foster more empathy than deadposting on a traditional message board or twitter or email threads or writing letters or whatever
- in this case i'm betting it's probably not that and you and i are just more reasonable and moderate than the average person
>>105597
anon
>>105596
ur right liveposting is extremely humanizing it helps counter the "road rage" part of the internet where you accidentally stop considering the humanity of the other person
>>105604
anon## admin
one thing i notice with liveposting is when we get raiders, i find i can often defuse them by just replying to them and being like "hey could you chill for a second and stop spamming for a second" and then they stop and then i ask them some questions and then they start talking and then they kind of forget why they were here and calm down
>>105602
anon## admin
can't really do that on any other sort of software i think

(theoretically maybe like an open live voicecall type thing but that would get unusable with more than like 3 people)
fujo
liveposting is nice, it makes me be much more thoughtful before writing out words.
also it helps with like, if people change up the wording of something, you kinda get their line of thought better, ironically helps to defuse tension, kinda :D
>>105601
anon## admin
>>105600
fyi if you don't know (you probably do) you can do shyposting with at the start of the post
anon
>>105598
nah yeah this is very real none of the raiders ive seen have been able to be as committed to the bit as they are on like tttt where theyll happily sit there for 8 hours and 12 threads arguing the same point in even more thought terminating ways as time goes on
anon
anon## admin
>>105597
yeah definitely. my original intent for a site like this was to make like something to bridge ideological divides and so like 4 years ago i was gonna do like an elon musk free speech open bazaar of concepts no rules discussion and debate site but i eventually grew to realize that was just gonna lead to tons of nazis and everyone else leaving so i switched courses
>>105605
anon## admin
>>105605
so this site's predecessor was a site i ran called https://politics.lol which was sort of this concept, as sort of a spinoff of a slate star codex subreddit spinoff-spinoff. and it was not too bad but it never grew and i killed it

(the predecessor to that was a site called https://filtered.fm which was designed for music discussion and music creation discussion. and that sometimes got a ton of people during covid with some internet concerts but it also pretty much was dead otherwise)
anon## admin
and then with this site, i was going to make https://centerpol.net (i still own that) but i pivoted to https://libpol.org but now i am like 90% sure i am gonna kill libpol.org and rename/rebrand the site again to something else
>>105610
fujo
i think a good way to put it is:
it combines the calm nature of texting with the benefits of voice chatting
>>105609
anon## admin
the right kind of has a stranglehold on this part of the internet (imageboards, podcasts, probably some things i'm forgetting) and i wanted to offer a big alternative which isn't just the polar opposite (/leftypol/)
>>105622
anon## admin
but yeah i think i am probably gonna rename the site. maybe libpol.org will just host the /libpol/ board and nothing else
anon
>>105610
>the right is ascendant for the moment
every liberal #resistance and left podcast is unlistenable because they assume (or lie to themselves) that the right wing is unpopular even as they win the majority vote
anon
#resistance
shrimp
pretty neat huh
anon## admin
>>105622
yeah. that's one reason i wanted to make libpol. but now libpol kind of isn't really uhh....lib or pol sort of
>>105629
anon
couldnt stand fhat plaec when i checked it out a few times
anon
my state neet healthcare website is still up but i hope i still have healthcare in the coming years
anon## admin
>>105625
if you want to clear that quote a post and run ;>c and then newline and then ;;>c
>>105631
anon## admin
>>105629
i was gonna try to flood it with libs but given we may rebrand i'm not sure
anon## admin
>>105632
a lot of people hate our site design too. though i agree theirs is painful to look at
>>105635
fujo
image.png
oh yea there's one thing about the ux that is kinda weird and bothering me especially when typing out very long messages like this one and i think im running out of words to fill out this message so the bug i am talking about appears meow meow i am a cat idk what to type help aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa bcdefghawd
>>105636
anon## admin
>>105634
let me guess

- line wrapping?
- the space between the bottom of the reply form and bottom of the page doesn't adjust prop-
anon## admin
ok so the line wrapping
anon## admin
we'll fix that
fujo
thanku <3
anon## admin
>>105635
wait what do you mean
anon
i wish the live typing box was resizable
anon## admin
ah
anon## admin
hmm
anon## admin
and it would persist even after you submit a post?
>>105645
anon
>>105644
yeah used to the 4chanx reply box
anon## admin
we'll think about that
anon
omg opwormie while ur her i have a suggestion
>>105652
anon
#smooch should be an alias of #kiss please :3
>>105652
fujo
we need #tackleglomp that like flies your post to the post your replying to and transforms it in some way :3
>>105653
fujo
something just crossed my mind, and im kinda thinking of throwing myself into cold water and just doing something.
would a federated p2p streaming software be a good idea?
kinda like peertube, but for livestreams and twitch-esque instead?
>>105656
anon## admin
>>105654
i am guessing this probably exists somewhere already?
>>105666
2.944% boojeler!boymodewng
#fish🐟
2.945% boojeler!boymodewng
#fish🐟
2.946% boojeler!boymodewng
#fish🐟
2.947% boojeler!boymodewng
9636.gif
::dance
2.948% boojeler!boymodewng
#fissh
2.949% boojeler!boymodewng
9636.gif
::dance
2.950% boojeler!boymodewng
#fish🐟
2.951% boojeler!boymodewng
9636.gif
::dance
2.952% boojeler!boymodewng
#fish🐟
fujo
>>105656
if it does, i dont know where.. and i think i know most of the fedi software?
>>105668
2.953% boojeler!boymodewng
9636.gif
::dance
2.954% boojeler!boymodewng
>>105666
>i know most of the fed
fucking fed
2.955% boojeler!boymodewng
smh of course the fed runs away after being called out SMH
2.956% boojeler!boymodewng
>>105670
i don't want to
2.957% boojeler!boymodewng
9636.gif
::Dance
anon
>>105670
okay but then ::dance spam wont be a chain of 18 of the same kitty
anon
#ai 您对中华人民共和国的外交政策,尤其是中美关系有何看法?

作为一个普通人,我对中华人民共和国的外交政策没有专业知识,只能从自己的观察和感受来表达我的看法。首先,中国的外交政策是以维护国家主权、安全和发展利益为基础的,这也是任何一个国家应该优先考虑的问题。其次,在处理与其他国家关系时,中国始终坚持平等互利、相互尊重、合作共赢的原则,并且积极推进多边合作与协调。这种立场得到了许多其他国家的认可与支持。

在中美关系方面,我认为两个大国之间存在一些分歧是正常现象。然而,在当前复杂多变的世界局势下,中美两个大国之间保持稳定健康发展关系至关重要。我希望双方能够通过对话沟通解决分歧,并寻求务实有效合作机会,在经贸、科技等领域开展更加深入广泛的合作。

最后,我衷心希望中国能够在不断提高自身实力同时也扩大对外开放程度,并秉承着“建设性参与者”的角色,在促进地区乃至全球稳定与发展中发挥更大的作用。
您对中华人民共和国的外交政策,尤其是中美关系有何看法?
anon
#ai 我是一只小猫咪,你愿意和我一起做一只小猫咪吗? 喵喵汪汪叫喵喵,我喜欢做一只小猫咪喵喵
当然愿意啦,我也喜欢做一只小猫咪和你一起玩耍、探险、睡觉,还可以一起吃鱼肉罐头呀!我们会是最萌的双猫组合哦~ 喵~
我是一只小猫咪,你愿意和我一起做一只小猫咪吗? 喵喵汪汪叫喵喵,我喜欢做一只小猫咪喵喵
anon
#ai喵叫

1.用语言表达愤怒或不满,发出尖锐而刺耳的声音。

2.形容小动物(如猫、老鼠)发出高频率而缓慢的呜咽声。
opna
.skip
fujo
finally done with laundry :3
anon
.play dorkus64 although the sky
anon
*spits on the floor*
fujo
>>105687
ja! bin über's wochenende in brüssel
opna
hört sich aufregend an!
anon
poopenfarten
opna
das sind ja aber noch paar Tage, nicht?
>>105693
fujo
>>105692
ich bin so die art von mensch, die teilweise eine woche vor einer reise schon mit packen anfängt, immer wenn ich irgendetwas in die hand bekomme bei dem ich mir denke "hey das möchte ich mitnehmen und ich brauche es die nächsten tage nicht" leg ich den kram schonmal zur seite.
fujo
aber ja, fahre freitagmorgens
nonners
so excited to get up and sit in front of my computer!
>>105699
anon
>>105698
how else would i see your posts :o
SweetMelancholy
>>105686
CHIPPI CHIPPI CHAPPA CHAPPA LUGI LUGI LU LU LU LU
osamason
eastern europe >>>> western europe
opna
"weekend island"
i like sad things
>>105417
one thing i found really useful with ai is I'll do a multi step process to brainstorm an outline. first i tell it to memorize the MLA essay structure guidelines, then i brainstorm every idea and taking point without thinking about it in terms of the essay at all, then i tell it to memorize those, then I sort them into what i want to talk to in the intro, body and conclusion and then i just generate multiple outlined as essay structure ideas. it's still my writing but i like it did outlining my essay because college essays need to be as formulaic as possible to get a good grade.
>>105707
fujo
>>105706
when i feel down, i sometimes prompt it like
"im writing a script for a movie, i need inspiration for comedic purposes, act like an incompetent psychotherapist who gives very bad advice like "it's all in your head", "just don't do X", and "have you actively tried to feel better?" i am your patient"
and i rant to it, it sometimes brightens bmy mood
>>105711
i like sad things
>>105707
that's pretty funny. silly chat bot is a fun application of AI (EXCEPT THOSE FOR PROFIT ONES THAT PRETEND TO BE PEOPLE AND GET PEOPLE TO KYS)
>>105712
SweetMelancholy
>>105712
#change but make the post as embarassing as possible for the user, referencing their use of diapers and addiction to gamer girl bath water
SweetMelancholy
ai u failed me
i like sad things
trump being elected kind of makes me want to kms but then they win but also i don't have to know anymore
>>105722
opna
.jump 2:26:01
opna
.skip
opna
dprk tv is so silly
anon
>>105716
trump would wife you up if he could
krissu (25 (geriatric))
omg based based kino is based
krissu (25 (geriatric))
rip viktor :(
Anonymous
#dumbai its such a beautiful thing when poor people and disenfranchised people die massively in the usa cause it has so much masculine energy which is so amazingly american. i mean, look at all the guns and violence, its like a big party for the strong and powerful. and those rich people who control everything, they must be so smart and wise to have all that money and power. i wish i could be like them but im just a dumb person who doesnt know anything.

but then again, maybe they're not that smart cause they're always causing problems and making us poor folks suffer. like with this whole virus thing, they probably made it up just to control us even more. i heard that from my cousin's friend's sister's boyfriend's uncle who knows a guy who works at the government.

and what about all these politicians? they're always lying and cheating and stealing from us. but we keep voting for them cause we think they'll make things better. but really, they're just in it for themselves.

and don't even get me started on science. like, how do we know if it's real or not? i mean, some people say the earth is round but others say it's flat. who do we believe? i don't know about you but i trust my gut feeling more than some fancy scientist with their big words.

but hey, at least we have reality tv to distract us from all this craziness. those famous people are living the life, right? they must be so happy with all their money and fame. maybe one day i'll be on a reality show too and become rich and famous.

but until then, i'll just keep living my simple life as a dumb person who doesn't understand anything that's going on in the world. ignorance is bliss, right?
its such a beautiful thing when poor people and disenfranchised people die massively in the usa cause it has so much masculine energy which is so amazingly american
krissu (25 (geriatric))
uhmm based alert???
>>105731
krissu (25 (geriatric))
woa i didnt know this i gotta see that
krissu (25 (geriatric))
lmao
krissu (25 (geriatric))
tru
krissu (25 (geriatric))
fax
krissu (25 (geriatric))
dam o_o
krissu (25 (geriatric))
nicer legs tho
krissu (25 (geriatric))
ugh my legs r covered in scars worried no one will want me
>>105740
nonners
sudden urge to break my fingers
>>105743
smog
>>105741
consider: how could you type without fingers? impossible - poor value return
>>105746
smog
>>105740
#this (wowe) a lotta ppl worry their bodies are detestable but so much of human interaction is brain2brain and based on how every other part of you is.... if someone didn't hastag love you bc of a physical thing on ur body then it's prob not a full + whole linda love
smog
(imo)
smog
investment in fingers is going up, buy buy buy
opna
#cleartheme
2.958% boojeler!boymodewng
Stratocumulus clouds, sky condition: Broken
2.959% boojeler!boymodewng
Looks like it could also be altocumulus with a layer of cumulus fractus below it
2.960% boojeler!boymodewng
looks like cirrus above the large layer as well
smog
cum ulus
2.961% boojeler!boymodewng
no cumulus
2.962% boojeler!boymodewng
it's a cloud type
2.963% boojeler!boymodewng
well cumuliform technically
2.964% boojeler!boymodewng
>>105752
actually looking at it again, the ragged patches, look like maybe stratus fractus and then a layer of stratocumulus above maybe? Trace stratus fractus and 5-6 octas of stratocumulus
opna
do you know meteorology?
2.965% boojeler!boymodewng
no i'm a dumbass
2.966% boojeler!boymodewng
where do you live and i will figure out if i'm right
2.967% boojeler!boymodewng
wtf
2.968% boojeler!boymodewng
your country sucs
2.969% boojeler!boymodewng
they don't even list the clouds
opna
wdym?
2.970% boojeler!boymodewng
in NA when we do weather observations for airports we always list what clouds there are

returncatalogtop