#I know I already posted a bunch and idk how algorithm works on here so this might flop

[norgeant]

"Ah yes. Me. My boyfriend. And his F1 contact for the next couple of years at least"

Max fewtrell can relate to this exact situation

[earlgreytea68]

Okay, I really came over here to post about artificial intelligence and not algorithms but those are one and the same, really, and illustrate the shortcomings of artificial intelligence.

I had a couple of people ask me what I was writing about with artificial intelligence and the truth is that I don't know. I actually find it difficult to even read artificial intelligence discourse because it all just sounds absolutely terrifying to me wrapped up in "if you find it terrifying, it's because you're old." But this is also my field and I feel like to be a responsible scholar I should have coherent thoughts about artificial intelligence, and so I keep trying to make myself learn more. But I don't know if I am at all sensible or logical about AI or just an absolute mess.

So, anyway, I'm in the process of brainstorming what my thoughts are. I am THINKING initially that my approach isn't really going to be "is this good or bad" or "is this legal or illegal," because I think a lot of people who understand the tech much better than me are having those debates. What really caught my attention is how whenever people talk about the ChatGPT type of artificial intelligence, what they say is, "This is just what people do! People read stuff and then they take what they've learned from that stuff and create new stuff!" And yeeeees, absolutely, that is exactly what people do, but when people do that stuff, like, when fans consume a bunch of stuff and then make something new out of the object of their affection, traditionally the legal system has been like, "ewwwww, what lazy hacks." So suddenly we've all discovered that this is how people work? And we're all cool with it now because we've taught some computers to do it? Are we still going to be cool with it when people like fans do it? Or no? Like, that's what I think I kind of want to point out and highlight? Do we let machines get away with things we wouldn't let humans get away with? Is a machine even "like" a human in the first place? Why are we framing this discussion this way? Idk.

Other thoughts I've had while researching:

People analyzing AI tend to treat it VERY technically. Like, "Well, obviously using all these works for input isn't copyright infringement because technically speaking no copies remain because they get broken down into data points," blah blah blah. It's just VERY fine-toothed-comb, this kind of analysis.

On the one hand, people talking about AI are like, "This is just what humans do!" And on the other hand, people talking about AI are like, "This is a revolutionary tool that will change the world!" And, well...which is it? If this is just what humans already do, why is it so revolutionary? Presumably because of the speed and size and scale of it, that THAT'S what's revolutionary. But I also think it's wrong to be like, "AI is just REALLY FAST human creativity," because AI knows everything, and no human does, and at the same time I cannot shake the idea that humans are different from machines, but maybe that's me being hopelessly naive. BUT I think it's a different type of naivete to be like, "Exactly, humans are different from machines, so machines will never replace them!" because...there is literally nothing about modern society that makes me think every industry everywhere won't jump on the ability to replace workers with machines. (This might conflict with my "make people miserable" economic theory except no, depriving people of jobs they want will definitely make people miserable, which is why I suspect AI will only replace the jobs people actually want to do.)

Copyright law mostly only works if you presume that everyone is out to monetize all creativity at all times. Some people are like, "We don't need to modify existing copyright law, because AI will fit into our copyright law model," so that must assume that AI will also be inevitably exploitative. Which is probably right, let's be honest, this is a capitalist society we live in. But allowing people to break out of exploitative models (reach audiences with the gatekeepers of traditional publishing / music labels / movie studios) has caused humanity to explode with incredible amounts of creativity. So being like, "Let's just use AI to exploit things again" seems weird to me. At the same time, though, much of AI stuff right now is not necessarily exploitative and still seems to be harming artists, so maybe this thought of mine is going nowhere. But I feel this little nibble of suspicion that we'll land somewhere where AI will be legal / accessible as a tool only to those paying for it and thus extracting a price for the output in order to justify the cost.

People keep saying that artists will learn to use AI so it will all be fine. Idk how I feel about the best outcome we can imagine being "creative people will adjust the way they create to accommodate new machines." Like, some people will learn to use AI and they will love it and it will be exactly how they want to create, and that's great! But some people will never want to create using AI and we should make sure both ways are okay but we are generally not good as a society at doing stuff like that, so. And when people say stuff like this, they never say, "Some artists will adjust," they say, "Artists will adjust," with presumably that implication that artists who don't will no longer be considered artists.

I keep coming back to: Why did we ever need AI for creative purposes? Like, I get why we need AI to be chatbots or do algorithms or go through predictive data or all this other stuff which suffers from the human limitation of not being able to be 24/7 calculating all the human knowledge in the world. I guess I just don't understand what about creativity humans weren't handling well enough on their own, that we needed computers to do it for us. I saw someone say that AI is not going to be used to make our lives better. This goes back to my theory of capitalism: AI will not be used to replace the drudgery thoughtless tasks nobody wants to do. AI will be used to replace the creative work that people actually desperately want to do. I don't see any reason not to think that's true, Idk.

I also understand that there's no way that I talk about AI and don't sound like a hopelessly old person upset about the invention of photography or something, so I really try really, really hard to view it as positively as I can. I think that AI is probably incredibly useful for people who struggle with expressing themselves through words and so AI can help them with that, or struggle with drawing and so AI can help with that. And so I don't want to discount those positive aspects of AI, either. (At the same time, I thought about using AI to help me draw stuff, since I can't draw, and found the idea deeply odd, but that's probably because the reason I can't draw is because I'm not a visual thinker, period, and for that reason I just never felt confident that what the AI gave me would be anything that would come out of my own head? If that makes sense?)

I just feel like we're standing here on the precipice developing this incredibly powerful tool and we have a terrible track record with basically any incredibly powerful tool we have ever developed sigh.

So yeah, no coherent thoughts. Idk if I ever will have coherent thoughts lol. I might end up abandoning the project.

[somnilogical]

they will never be as strong or as fast as i can be

copy/pasted from a convo:

<<somni: ive been exploiting being able to talk about everything vs miri/cfar cant do what i do bc if they did they would talk about how they are evil. it would all chain back.

somni: omg i can just post this to my blog because i can talk about my meta-strategy and it confers pretty much no relative advantage to miri/cfar. because 1 most of them have disassembled their agency so its like talking in front someone who works at the dmv about taking over the world and the ones that have any agency (basically just anna salamon) have to work with and coordinate via brokenness the masses that have and 2 feels secure in the way that saying ill use my soul as my weapon feels secure, like the power of this technique doesnt depend much on people not knowing im using it.>>

truth is entangled and lies contagious. justice is entangled and injustice contagious. in order to sustain their facade, miri/cfar had to chain back to lie about the principles of decision theory itself. lie about the organization structure of cfar, lie about miri's fundraiser. and so much more.

any series of reasoned claims they make will chain back to stuff thats false or injustice, because they seek to maintain a region of untruth and injustice.

so yeah, miri/cfar basically cant talk in public except in staid formalities infinitely pouring the same entropy of "these people are psychotic" "these people are infohazards" "do not read what they write" "stay the course" "everything is under control, do not panic" "i know my associates at miri/cfar, they are good people" "if you talk with these people you may become a rapist". but not actually able to manifest dynamic compute. to explain themselves they built their own personal room 101, filled with miri/cfar affiliates and formed a united front of gaslighting. deluks (author of that one rationalist blog where they worked to read and summarize all the others) talks about the kind of compute miri/cfar manifested:

<<deluks: I also updated a lot based on Bay Area safety discussion

idk if I have ever been in such a hostile environment for anyone trying to discuss making thigns safer

If you wanted to discuss how Anna et all were innocent people would happily chat with you

If you tried to discuss ideas for making things safer either you got silence

or people would be insanely hostle if you plausibly slipped up at all

or even seemed like you might have been not careful enough in how you phrased things

extremely careful -> no engagement at all//even slightly less care -> get dogpilled>>

they have picked up the optimization style of of cops, as alice maz described them:

<<the role of the cop is to defend society against the members of society. police officers are trivially cops. firefighters and paramedics, despite similar aesthetic trappings, are emphatically not. bureaucrats and prosecutors are cops, as are the worst judges, though the best are not. schoolteachers and therapists are almost always cops; this is a great crime, as they present themselves to the young and the vulnerable as their friends, only to turn on them should they violate one of their profession's many taboos. soldiers and parents need not be cops, but the former may be used as such, and the latter seem frighteningly eager to enlist. the cop is the enemy of passion and the enemy of freedom, never forget this>>

i can travel lots of places and regenerate truth and justice.

i can go to a trans support group in the bay and show them logs of what elle said and did and they can recognize the pattern of minority oppression, transmisogyny.

i can talk with uninvolved decision-theorists about why paying out to oneshot blackmail with subjunctive dependence because "In game theory, paying out to blackmail is bad, because it creates an incentive for more future blackmail." is wrong. and why exploiting your subjunctive dependence as a udt agent to not pay out is right. they cant.

--

miri/cfar have to centrally coordinate on lies or they start crashing into each other. independently generating falsehoods in isolation makes them point in all directions.

independently generating and working off of truths allows everything to point in the same direction without needing to communicate. i can write this post and then idk maybe someone im algorithmically colluding with on this writes another post and they dont come out all distorted and skew with each other. this caches out in what looks from the outside as an uncanny ability to start dynamically colluding with people and output distinct strains of philosophy based on shared precepts.

interference with yourself looks like kelsey piper trying to claim that emma and somni are starting some sort of rape cult and anna and miri/cfar trying to claim we are naive victims of ziz's cult and ▘▕▜▋ claiming emma and somni are mindhacking ziz to make her bully them and jade nameless claiming im doing this to get a job at cfar and ...

since they make up their fake coordination points independently they smash into each other. if they want to coordinate over lots of people they then have to work out which of these they want to coordinate around in a sort of market of falsehoods. and have to arrange for it to not contradict any information anything people know. but they dont know all the information everyone knows, and they wont know it even after combing through lots of blogs and reading lots of discord chats.

when they try coordinating on falsehoods like this, its hard to get a coalition together in an environment where what people know is rapidly changing because a bunch of anarchist bloggers keep posting things in a bunch of places on a non-centrally controlled schedule determined by what seems like a good idea at the time to independent agents. and having lots of conversations with so many different people in private and public they cant keep track of them all.

if they try pretending to be dumb and forming a unified gaslighting front in one area. then people will exploit the fact that this is the internet and not the evolutionary environment, take logs and post them somewhere else where everyone didnt collude to be dumb in this particular way. so while their monkey brains get a rush of endorphins from being able to successfully coordinate local humans, what feels like an entire tribe, against the blasphemer, actually they just used their adult intelligence to defeat in front of a bunch of people who dont share their political commitments but who can reason about what is true and what is just.

(of course there are many truths this doesnt work on because of large inferential distance, shared mammalian biases it takes an unusual mind to step over, and shared incentives. but the defense of most regions of injustice and untruth when you ask questions have to keep chaining to more and more absurd things until you are defending causal decision theory or start claiming 'anna salamon, the president of cfar, is not involved in cfar's hiring'. which depend on a social context committed to defending everything that protects miri/cfar and people who dont have the same conclusion-that-must-not-happen can see that its dumb.)

if miri/cfar had committed themselves to the path of expanding agency, maybe i wouldnt be posting my thoughts and meta-process on the public internet. (in the counterfactual where they committed to this path, its likely that i wouldnt be protesting. because it seems actually-hard to stay on the path and remain evil.) but as it stands, i expect this information to differentially help anarchists and do about as much good for statists as explaining updateless decision theory to someone at cfar. its just this inert structure in their brains, they cant do anything strategic with it. they intentionally shut down their ability to take ideas seriously and drive out anyone left who can, calling them crazy.

what they can do is "oh here is a list of people to target" and "see if they said anything incriminating". ive seen their attempts to coordinate enter the attractors of 'authoritarianism' (duncans dragon army, kingsleys "repent and submit to [AUTHORITY FIGURE]") and 'lets all lie in the same direction and disable general cognition to update out of this! the important part is social agreement and that everyone allows social reality to have the final veto on their beliefs. i myself do this so you know im super safe and this is super fair.' (anna and kelsey). this sort of weak coordination based on breaking people can be easily subverted by anything real.

--

if you are actually right, you can exploit useful properties of being right and let that be your asymmetric weapon. such that all that challenge you know they will know its steel. and then people who compute the outcome and expect to lose, dont fight in the first place.

if my chosen weapon were actually the size of my muscles and imposing figure compared to anna salamon as miri/cfar people "believed" (exploiting the already extant anti-transfem psychic suppression field as one of their few functioning coordination points. probably not as functional now after what i have written.), then when i fought people it would create a warp field such that then people with smaller muscles wont fight in the first place, but id be deluged by people with larger muscles. i dont want to create a warp field that summons people with lots of muscles.

if i exploit properties of my souls, of truth and justice. then i have an arsenal of techniques that are stronger if i actually want to save everyone, if im actually right, if im acting for justice. because they exploit useful differential properties of each. and the warp field in higher density summons ... people who care about saving the world, truth, and justice. in other words, a high density of potential allies.

by default i want to exploit "the difference is that im right" not "the difference is that i have larger muscles". i want differential power to push away those who are wrong and unjust and attract those who are right and just into a kind of warp hull.

there are other reasons as well.

[thelightofthingshopedfor]

have 8 million things to do very soon and I can tell that I'm riiiiiight on the edge of a depression spiral (which is itself at least partly caused by being overwhelmed, dealing with it poorly and becoming more overwhelmed, and hating myself for dealing with it poorly) which is SUPER GREAT TIMING. so obviously I'm making a list, because sometimes that helps.

things I absolutely must do this weekend:

finish the hair repaint on my current Etsy order so I can mail it Monday or earlier

do anything I might need to do for the other outstanding order that the customer also wants in time for Christmas somehow, so that when the parts arrive on MONDAY I can put the thing together and mail it almost immediately (damn well better show up on Monday, which is itself irritating because the site said free 2-day shipping but nobody really means that when they ship to Alaska)

wrap and prepare any gifts I want to mail out because I should also send those no later than Monday, and in fact earlier would have been better, but for some reason every year I'm like "eh, it's fine, there's time" right up until there suddenly isn't

do...whatever I still need to do with Tumblr. import to Wordpress, back up the whole thing to my computer with one of several methods I've reblogged, try to find specific posts I want to save (original posts and anything tagged "fic ideas" will be tedious but at least easy to find; no idea about other stuff)

semi-related: add something to my Dreamwidth profile so it's not...blank

haul some more stuff to Value Village and other donation sites, because I’m about to pick up a bunch of shit that will take up too much space in my car (somewhat related: books, music, movies, and toys are 40% off today and tomorrow, which is unusual for them)

set up the damn tree, like even if we don't put ornaments on it I'd at least like to have the tree up (requires digging it out of the garage, which is a disaster and very much not my disaster)

do something with my dad and sister for my birthday (but first, figure out what and when, and like...I don’t have the mental energy for that)

try to get more birthday donations for SPLC (I went with Trevor Project last year and got several donations without doing much to promote it, so I don't know if this one's getting a lot less because it seems more political, or algorithms are hiding this and a lot of my other posts, or a lot of people have deliberately snoozed me because of my political posts, or...something else, idk)

transfer stuff into my new planner, because my current one is close enough to the end that it's not very useful anymore

other things that don't necessarily have to happen this weekend but should happen in the really near future, both because they need to get done and because they're contributing to my overall mental load that means everything is overwhelming:

figure out how to send in the claim for my car accident last year to my life-insurance-and-a-few-other-things company, because it's worth $50 (should probably also see if I can get something similar for the earthquake, because I think it did fuck up my neck a little more, and $50 is $50 when I'm paying them that much each month)

list other things on Etsy...now that it's too late to take advantage of holiday sales 😖

sign up for actual training with Hazy

also, bug the rescue group again about recent vet info to figure out whether she has any pre-existing conditions and when she needs a checkup (maybe soon, because she seems to scratch herself a LOT and sometimes she gets kind of wheezy)

also also, try to figure out ways to work on her separation anxiety and general hyperness, which I DON'T KNOW HOW TO DO because it's all a gradual process that I don't really have time for, and I have yet to find a smart/puzzle toy or a good chew toy that really seems to occupy her attention so I guess I still need to try more, and apparently mental stimulation through training can be good for anxiety and hyperactivity both so I should really be working on extremely basic commands on my own probably??

shit that's right I also need to figure out if I want to get her on VCA Care Club and do research for pet insurance that might be better than the one we had for Scully

get myself back into a better Planet Fitness habit, because I didn't go at all this week and I've definitely dropped off in the last few weeks, which is partly because I've been busy with other things INCLUDING THE GIANT EARTHQUAKE but partly because the main reason I was good about it before was that I literally tricked myself into wanting to work out so I could listen to The Adventure Zone, and once I caught up on that, nothing else I've tried (MBMBAM, occasional TAZ updates, Night Vale) has quite filled the gap of "hilarious goofs + suspenseful long-form narrative" that makes me actually want to work out, so...either I need to figure out how make myself go without that motivation (mixed success so far or I wouldn't be talking about it or feeling guilty about it) or find another podcast I want to binge. and try to figure out if there's any point to asking them to stop playing so many news channels (especially Fox two or three times and fucking OAN once) at the TV bank for the cardio machines, because that is...honestly another thing that makes me not want to go

related: somehow find a way to make myself do my core exercises regularly, because slacking on those and slacking on Planet Fitness is probably part of the reason my headaches and neck and general constant exhaustion have maybe been worse lately, AND ALSO start regularly doing the exercises that are supposed to help with my newish hand/wrist pain, all of which sucks because a big ol' theme here is me having a hard time forming un-fun new long-term habits (exercise, going to bed earlier) that are crucial for improving basically every part of my life and not worsening the health problems i already have

TYPE UP MY DAMN NOTEBOOKS and organize everything so I actually know how much I have in my current WIPs; now that my computer is back there's no excuse for not doing this

also like...write. in general. now that I have a Christmas-related idea I'd kinda like to do and I also want to do a Yuletide treat, and maybe Avengers: Endgame isn't really a deadline for lots of other fics but also it kind of is

actually organize my backups so they're not a disaster and it's not a crisis next time I have computer problems (plus like, I'm going to need everything backed up when I upgrade)

research and buy a CPU, motherboard, SSD, and maybe new PSU, ideally without spending a horrendous amount of money

make more progress in SWTOR because there's no guarantee how long it's going to be around and tbh it's ridiculous I haven't finished all the class storylines despite having been a subscriber for like. SEVEN YEARS

unfuck my iTunes library YET AGAIN

find a therapist, because I've probably been needing one for a while but a huge theme in all of this is being overwhelmed because I have too much to do and don't know how to deal with it, and then getting into guilt and self-loathing because I dealt with it badly and it got worse, and I'm increasingly sure it's my not-really-diagnosed-or-treated ADHD starting this old, old cycle to begin with and that means it's even more important to find a therapist who will actually. do stuff. with the ADHD. instead of just kinda...dropping it.

slight problem though, adding yet another regular appointment means less time for...everything else and that doesn't exactly help with being overwhelmed, so it's like...a disincentive to pursue it

the only thing I really want to do:

sleep for about a year

[robitherat]

I already said this all in the notes but I'm gonna rb with it too because I can and pandora said they were okay with it <3

I hate hate hate dictating how other people post so much because like. I get it! A lot of people's interpretations of masky and hoodie even as creepypasta characters draws from or even correlates with marble hornets. Like there's definitely gray areas in it and if mh is relevant even just in the way you interpret masky and hoodie, or if you include like Alex and Jay in your interpretations that's cool! Marble hornets and the related tags are relevant there! Absolutely use them! I'm not into creepypastas as anything more than like, idk, looking st what people do with the classic characters out of nostalgia or something, but I do sometimes like to see the overlap between marble hornets and crp! It can be a lot of fun and even if it's not for me it is for other people and I like seeing people enjoying things

The thing that really bugs me about it is this: tumblr's tag system is for organization. Its not like Twitter or Instagram or tiktok where ypur tags affect your reach. Tags are for tagging relevant fandoms so people in them can see your posts, tagging triggers for blacklisting, blog organization, but not reach. So when you add marble hornets tags to a post for creepypasta, and ESPECIALLY posts that don't even mention masky or hoodie in ANY way shape or form, you are actually spamming the tag according to tumblr TOS. Like, people can report you for it, and it would have the potential to get your post taken down.

Personally, it's just a major bummer to go into a tag specifically to see content for something I enjoy, and instead be met with a stream of posts for characters I either don't care about, or don't even recognize because they aren't actually relevant to that tag

If it were just a couple posts here and there it'd be one thing, but this is a CONSTANT issue. And it's not just here! Plenty of other fandoms and tags have the same issue. Hell I remember discourse in the Ethan Winters tag over one specific blogger tagging a post saying howuch they hate his character in the Ethan winters tag. And then they got upset because people were pointing out how stupid and shitty that was. I didn't see all of it, I'm sure some people got way too into it and were also being shitty, but point is: using the relevant tags and ONLY the relevant tags is just basic etiquette. If you're saying how much you hate a character, don't put that character's tag on your post, because then people who like that character will see it, they'll get mad at you because they go to the tag to see stuff about a character they LIKE, you get mad because you have a whole bunch of people getting mad at you, everyone loses. If your post has nothing to do with a fandom that's maybe sometimes adjacent, don't tag it, because you're not actually getting any more notable interaction than you would have normally.

Basically, we need to keep in mind and continue to make sure others know that Tumblr doesn't work like other sites, and that goes especially for the tag system. There's no algorithm, tags aren't for reach, stop fuckin with the system. It'll make the ride way smoother for everyone. PLUS, it means I won't have to block you for being annoying and clogging up the tags.

If people want to look at your posts, they will. Trust me, if you post enough your people will find you. You've just got to post in the right tags so they can.

i also wanna say, i get a lot of the crosstagged creepypasta marble hornets posts comes from a place of ‘oh, thats where xyz character originated from, ill tag is so fans of that character can see!’ 

but in my experience as someone whos been into mh actively since around 2015, (which, considering mh started in 2009, isnt actually all that long) the majority of mh fans want to keep marble hornets and creepypasta separate. and a lot of us are creepypasta enjoyers! but when i wanna see content for, like, tim wright, i dont also wanna see toby and jeff the killer. its not the content im trying to find.

yes i can blacklist certain tags so they dont show up for me but the point of the matter is that for the most part, when you go into a fandom tag on tumblr, you expect to see only content for that fandom unless someone is explicitly posting about their crossover au (which is totally fine and a totally diff situation to this)

but i got into the mh tag nowadays and its like… majority creepypasta posting thats been crosstagged. yall are… youre aware you can post about your creepypasta versions of masky and hoodie without tagging marble hornets right. like its what everyone did for years before this. 

maybe i missed a memo or smth but this is kinda ridiculous as this point