Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Rappaport
Oct 2, 2013

SubG posted:

So my suggestion is that perhaps the reason why intelligence seems to be a jumbled mess of disjointed ideas is because that's precisely what it is.

Of course it is. I believe I've used several iterations of parables at this point to say that it, intelligence, can't be just measured neatly. But my point is, the colour of a beer bottle is something that can be altered (or measured, precisely), but there isn't a clear-cut way of making a student learn something they have a hard time with, for whatever reason. The entire field of pedagogy can be thrown out the window the minute humanity figures out a way of "painting" people capable of anything, but insofar as I know that has not happened yet. Kids take uppers to cram for exams, sure, but that doesn't make any (random sample) of them Paul Erdös, who also took uppers because he found numbers instead of beer bottles delightful. I can't say for sure whether my pocket calculator is more intelligent than Erdös when it comes to figuring out numbers, but my pocket calculator is less intelligent than David Lynch when it comes to making television entertainment. And so on. It is both a function of human social interactions, and a thing that is inherent to the person, and it seems silly to me to jump to the conclusion that intelligence, as a concept, cannot exist because we cannot just use a geiger counter to read off someone's intelligence score like in Fallout.

I am far more amenable to the idea that consciousness is ultimately simply a social contract, and a function of our mammal brains interpreting things like faces and social cues, which makes it harder to say whether Skynet is conscious in the same way as a cat is, and maybe the concept is useless when discussing alien minds. That's fine too.

Adbot
ADBOT LOVES YOU

Bar Ran Dun
Jan 22, 2006




Tei posted:

Do you reconoce your own voice on a record ?

I have my doubts.


All my ability to use language was learned by feedback from people using language around me.

Tei posted:

First, we can build weird poo poo, not just antropomized stuff.
It can be informed from something else than us.
Self-learning machines are a black box.
It could be like us, but primitive, rendering it alien.
it could be like us, but like a person with asperger or autism, rendering it alien to most poeple.
It could be build by a AI, and not us. Maybe they would talk in a made up language that only them understand.
Will not have human experiences.
Will not have a human body.
Will not have a human lifespan.
Will not have human needs.
Will not have human instincts.
It may be more rational than what humans are acustomed, making talking with him hard the way sometimes is hard to talk with programmers and other logical people.
It could just be designed around a different design than us. Making it very different to us in a very deep way.

All I’m seeing here is coming up with a constructed language to teach it and choosing it’s senses and then the problem of translation. It’s harder by degree but not in kind. It would have a different umwelten, but we would have selected and shaped that that unwelt.

SubG
Aug 19, 2004

It's a hard world for little things.

Rappaport posted:

It is both a function of human social interactions, and a thing that is inherent to the person[...]
That's your conclusion, but I really don't see how you're reaching it from what you enumerate before. You give a short list of human behaviours. They're loosely related in that they're things that most people would probably associate with "intelligence" (although in a fairly uneven and haphazard way—making a TV show might be evidence for "intelligence", but being unable to make a TV show isn't evidence for the absence of, or the presence of less, "intelligence"). But okay. My position is that "intelligence" seems to be a label we apply to a vague constellation of human behaviours instead being a description of a process or property inherent in objects. Your argument against this is...to list a couple of random human behaviours. Unless you're trying to argue that the "kids" in your example, or Erdös or Lynch, aren't human&mdashlor the things they do aren't human behaviours—I'm not sure I see the actual argument.

Rappaport
Oct 2, 2013

SubG posted:

My position is that "intelligence" seems to be a label we apply to a vague constellation of human behaviours instead being a description of a process or property inherent in objects. Your argument against this is...to list a couple of random human behaviours. Unless you're trying to argue that the "kids" in your example, or Erdös or Lynch, aren't human&mdashlor the things they do aren't human behaviours—I'm not sure I see the actual argument.

A beer bottle's colour can be ascertained by examining it, and intelligence may also only be ascertained by observing it "in action", if you will. I listed a couple of random humans, because it is evident that someone like Erdös was very mathematically talented, intelligent in that respect, and I picked Lynch for my artist example but we can take someone like da Vinci instead, he was intelligent in a wide range of things. If it isn't clear to the argument, the idea is that Leonardo (not the turtle) was inherently more gifted, intelligent, than someone I could randomly pick off the street. Intelligence, while being a vague constellation if you so insist, still appears to be a clearly and demonstrably physical aspect of humans, or other animals. Again I point out to the lobotomy example, my abilities in the vague constellation of intelligences would be reduced were my brain matter reduced surgically. For some definition of surgery, at any rate. This does not seem an especially controversial statement, because people have been lobotomized, and people have also suffered from other sorts of injury or illness that affected their brain in some way, reducing their demonstrable capabilities in the vague constellation. How is this not evidence of the vague constellation being a physical, inherent property of the physical creatures themselves?

The trouble with the AI example and consciousness is precisely that it cannot be measured like a beer bottle's chemical composition can be. But its intelligence is still a sensible thing to discuss, even if it is "artificial" in the mammalian sense or distinction. Does consciousness, or intelligence, need to mimic that which appears familiar to us, who already possess an intelligence and (arguably) a consciousness?

Rappaport fucked around with this message at 06:01 on Jul 15, 2023

Alctel
Jan 16, 2004

I love snails


Buncha people who've read Blindsight/Echopraxia ITT

SubG
Aug 19, 2004

It's a hard world for little things.

Rappaport posted:

A beer bottle's colour can be ascertained by examining it, and intelligence may also only be ascertained by observing it "in action", if you will. I listed a couple of random humans, because it is evident that someone like Erdös was very mathematically talented, intelligent in that respect, and I picked Lynch for my artist example but we can take someone like da Vinci instead, he was intelligent in a wide range of things. If it isn't clear to the argument, the idea is that Leonardo (not the turtle) was inherently more gifted, intelligent, than someone I could randomly pick off the street. Intelligence, while being a vague constellation if you so insist, still appears to be a clearly and demonstrably physical aspect of humans, or other animals. Again I point out to the lobotomy example, my abilities in the vague constellation of intelligences would be reduced were my brain matter reduced surgically. For some definition of surgery, at any rate. This does not seem an especially controversial statement, because people have been lobotomized, and people have also suffered from other sorts of injury or illness that affected their brain in some way, reducing their demonstrable capabilities in the vague constellation. How is this not evidence of the vague constellation being a physical, inherent property of the physical creatures themselves?
You appear to just be equivocating between "ability to do mathematics" and "intelligence", when we've already established they're not synonymous: "intelligence" encompasses other things (social awareness, rule induction, linguistic ability, and so on); and "ability to do mathematics" is neither necessary nor sufficient for "intelligence". Not necessary: presumably you don't think people who aren't Erdös (or people without Erdös numbers, or even people who can't do math at all) aren't intelligent, and we are willing to rank "intelligence" such that a human is smarter than a dog who is smarter than a bottle-loving beetle, and we're not doing the ranking by mathematical ability. Not sufficient: a pocket calculator can to mathematics but we seem to be in agreement that it's not "intelligent".

So if you want to argue that Erdös' ability to do mathematics was at least partially attributable to his physical characteristics, sure. In the same way that a pocket calculator's ability to do mathematics is a consequence of its physical characteristics. But that's not "intelligence". Which is why this is a discussion instead of just looking up "intelligence" on wikipedia to find out "ability to do mathematics".

(USER WAS PUT ON PROBATION FOR THIS POST)

Rappaport
Oct 2, 2013

SubG posted:

You appear to just be equivocating between "ability to do mathematics" and "intelligence", when we've already established they're not synonymous: "intelligence" encompasses other things (social awareness, rule induction, linguistic ability, and so on); and "ability to do mathematics" is neither necessary nor sufficient for "intelligence". Not necessary: presumably you don't think people who aren't Erdös (or people without Erdös numbers, or even people who can't do math at all) aren't intelligent, and we are willing to rank "intelligence" such that a human is smarter than a dog who is smarter than a bottle-loving beetle, and we're not doing the ranking by mathematical ability. Not sufficient: a pocket calculator can to mathematics but we seem to be in agreement that it's not "intelligent".

So if you want to argue that Erdös' ability to do mathematics was at least partially attributable to his physical characteristics, sure. In the same way that a pocket calculator's ability to do mathematics is a consequence of its physical characteristics. But that's not "intelligence". Which is why this is a discussion instead of just looking up "intelligence" on wikipedia to find out "ability to do mathematics".

You are quoting a post where I referenced Leonardo (not the turtle), too, and he was talented and intelligent in ways mathematical but also not mathematical. This entire response seems like a fairly weird "gotcha". I do not believe intelligence is readable on a person like a Fallout S.P.E.C.I.A.L. number, but that does not invalidate the idea of intelligence as a concept.

Rappaport fucked around with this message at 09:29 on Jul 15, 2023

SCheeseman
Apr 23, 2003

General intelligence is a hotly debated concept given it's origins lie in eugenics, it seems like a great way to marginalize life (though particularly the disabled and uneducated). "Intelligence" it's just too broad, it means too many things, which is why conversations about it end up like this. Measure intelligence by discreet skills and computers are intelligent, start measuring things and adding them together and you steer towards generalization again.

SCheeseman fucked around with this message at 10:19 on Jul 15, 2023

Serotoning
Sep 14, 2010

D&D: HASBARA SQUAD
HANG 'EM HIGH


We're fighting human animals and we act accordingly

SubG posted:

That's your conclusion, but I really don't see how you're reaching it from what you enumerate before. You give a short list of human behaviours. They're loosely related in that they're things that most people would probably associate with "intelligence" (although in a fairly uneven and haphazard way—making a TV show might be evidence for "intelligence", but being unable to make a TV show isn't evidence for the absence of, or the presence of less, "intelligence"). But okay. My position is that "intelligence" seems to be a label we apply to a vague constellation of human behaviours instead being a description of a process or property inherent in objects. Your argument against this is...to list a couple of random human behaviours. Unless you're trying to argue that the "kids" in your example, or Erdös or Lynch, aren't human&mdashlor the things they do aren't human behaviours—I'm not sure I see the actual argument.

idk what Rappaport would say about this but it seems pretty obvious to me that intelligence in the abstract is a process borne of a property (i.e., an ability), possessable by both humans and animals, and that many human behaviours are instances of intelligence. Doing mathematics and making movies require intelligence, manifest in these examples as an ability to reason logically and appropriately capture a story in video form, respectively.

SubG posted:

You're saying there's something that goes on inside objects to which intelligence is imputed independent on the social construct being used to evaluate them. Sure. Just like a beer bottle has a colour and texture independent of whether or not a beetle is currently trying to hump it. Whatever property you're observing in the object imputed to be "intelligent"...ability to do math or sensitivity to social cues or capacity for language or abstract reasoning or whatever...isn't inherently "intelligence" any more than "brown and shiny" is inherently "fuckable". Or at least it certainly doesn't seem to be. McCarthy et al coined the term "artificial intelligence" in the '50s and Searle came up with the Chinese Room in the '80s and oceans of ink have been spilled on the nature of "intelligence" and there's still no consensus. It is of course always possible that there really is some simple core concept that links together the jumbled mess of things that we associate with "intelligence", but if there is one we haven't been able to find it, and it's not for lack of trying. So my suggestion is that perhaps the reason why intelligence seems to be a jumbled mess of disjointed ideas is because that's precisely what it is.

I see now where your and Rappaport's quibble comes from. You are insistent to not merge/use the two most common definitions of intelligence, which include both the capacity and quality of acting intelligently, while Rappaport is fine with doing so. See the first two definitions here https://en.wiktionary.org/wiki/intelligence. I agree with Rappaport in that the capacity to do things associated with acting intelligently -- deduction, induction, etc. -- is itself intelligence, and so yes, that is an inherent property of an entity. To the extent that my toddler can use tools available to him to get to the snacks he wants in the kitchen, communicate the things he wants to, understand what a toy is doing, etc., he is intelligent.

Serotoning fucked around with this message at 10:45 on Jul 15, 2023

SubG
Aug 19, 2004

It's a hard world for little things.

Pvt. Parts posted:

You are insistent to not merge/use the two most common definitions of intelligence, which include both the capacity and quality of acting intelligently, while Rappaport is fine with doing so.
No, I'm observing that the capacity and the quality of doing various things (like engaging in mathematics, using language, or whatever) are unified as a single concept, "intelligence", primarily because of human social convention.

The Rio Grande is a river. At any given point it might have a riverbank on one side and another on the other side. A river having, and separating, riverbanks is a fundamental property of rivers. But now if you point at one riverbank and say it's Mexico and the other and say it's America, that's a social convention. There's no inherent property of one riverbank that defines its "Mexico-ness" or "America-ness". The social convention extends miles in either direction, assigning bits of geography to one category or the other. Many of them follow natural boundaries, but the categorisation is entirely the product of social convention.

My point is that "intelligence" is a social convention like that. Its definition in some places traces along what look like natural boundaries—there's no denying that ability to use language is a real, observable thing with real, observable consequences, just like a one riverbank being across the river from the other is real and observable—but grabbing that particular set of things and drawing a boundary around it is purely a social convention. And then we don't even follow that. Because ChatGPT can use language, and I think we all agree it's not "intelligent".

And when we decide okay, mathematics is definitely a sign of "intelligence"...then we end up having to do a bunch of tap-dancing around to explain no, not like pocket calculators, doing math like that doesn't count. Because, for example, the calculator can't decide to do math on its own. So what does it mean for a person to decide to do something on their own? Are people sources of acausal actions? Well [insert last half century or so of debate on the subject]....

And so on. It's possible, as I've said, that all this fuzzy and imprecise language that makes it look like "intelligence" just means "resembles human behaviour in some specific ways" is a just a coincidence and actually there's some single, concrete thing (an "intelligence" algorithm or something) underlying it all. But the position I've been advocating is that it looks like a fuzzy and imprecise term for a collection of various human behaviours because that is precisely what it is.

Reveilled
Apr 19, 2007

Take up your rifles
In this analogy, if “Mexico” and “America” are the things you call intelligence, what’s the thing that’s analogous to the riverbank?

Serotoning
Sep 14, 2010

D&D: HASBARA SQUAD
HANG 'EM HIGH


We're fighting human animals and we act accordingly

SubG posted:

No, I'm observing that the capacity and the quality of doing various things (like engaging in mathematics, using language, or whatever) are unified as a single concept, "intelligence", primarily because of human social convention.

The Rio Grande is a river. At any given point it might have a riverbank on one side and another on the other side. A river having, and separating, riverbanks is a fundamental property of rivers. But now if you point at one riverbank and say it's Mexico and the other and say it's America, that's a social convention. There's no inherent property of one riverbank that defines its "Mexico-ness" or "America-ness". The social convention extends miles in either direction, assigning bits of geography to one category or the other. Many of them follow natural boundaries, but the categorisation is entirely the product of social convention.

My point is that "intelligence" is a social convention like that. Its definition in some places traces along what look like natural boundaries—there's no denying that ability to use language is a real, observable thing with real, observable consequences, just like a one riverbank being across the river from the other is real and observable—but grabbing that particular set of things and drawing a boundary around it is purely a social convention. And then we don't even follow that. Because ChatGPT can use language, and I think we all agree it's not "intelligent".

And when we decide okay, mathematics is definitely a sign of "intelligence"...then we end up having to do a bunch of tap-dancing around to explain no, not like pocket calculators, doing math like that doesn't count. Because, for example, the calculator can't decide to do math on its own. So what does it mean for a person to decide to do something on their own? Are people sources of acausal actions? Well [insert last half century or so of debate on the subject]....

And so on. It's possible, as I've said, that all this fuzzy and imprecise language that makes it look like "intelligence" just means "resembles human behaviour in some specific ways" is a just a coincidence and actually there's some single, concrete thing (an "intelligence" algorithm or something) underlying it all. But the position I've been advocating is that it looks like a fuzzy and imprecise term for a collection of various human behaviours because that is precisely what it is.

Who said anything about human behaviours? I can't tell if you are being overly pedantic or overly human-centric, or both. Intelligence exists outside of humans and humans behaviours, and we know this very well already because we have animals, who display what we call intelligence. When a crow solves a puzzle or a monkey uses a tool, it is utilizing intelligence, safe from view of human social whatever. You are the one unifying intelligence under some weird human umbrella, no one else is, except maybe for the case of common parlance but even then, this is technical discussion and we don't need to do that here.

What I think you are getting at, and what I think is a much more fruitful talking point than endlessly trying to draw the correct shaped box around "intelligence", is: what does it mean for something to be intelligent if it lacks a conscious? I don't like your calculator example. What a calculator does is crunch numbers, which is "math" the same way spelling words correctly is "writing". Math is the language of logic and abstraction, and successfully creating or discovering it (depending on your view) requires intelligence more or less by definition. ChatGPT is a more interesting case because it appears to use language to communicate facts, knowledge, art, and so on. Surely, it must be "intelligent"? But does it "understand" what it writes in the way that a monkey understands what it is doing when it uses a stick to pick out an ant from a log? That is, it does have an internal model of the world from which it draws knowledge required to carry out a task? It could be said that a neural net is such a kind of model.

SubG
Aug 19, 2004

It's a hard world for little things.

Reveilled posted:

In this analogy, if “Mexico” and “America” are the things you call intelligence, what’s the thing that’s analogous to the riverbank?
I'm not sure if I'm understanding what you're asking, but in the analogy one riverbank would be something like "can use language" and the other would be "cannot use language" (or an analogous construction like "can do mathematics"), and the river would be whatever inherent physical characteristics divide the two. You can point at the river and say that an inherent difference (or some number of them) separates things that can use language and those that can't. But if you then say that one side is "intelligence" and one is "not intelligence", you're no longer talking about the inherent difference, you're talking about something else (which I'm arguing is "just" social convention). In the same way the Rio Grande observably has a north bank and a south bank, but those are only "America" and "Mexico" by convention. "America" doesn't just mean "north of the Rio Grande". You couldn't draw an accurate map by just using that as a guide, for example. Similarly, "can use language" does not appear to be an adequate definition of "intelligence", taken either intensively or extensively.

The obvious approach would then be to define the boundary as a collection of things like this: that is, say that "intelligence" is the region defined by something like "can use language AND can do math" (extending this to encompass however many characteristics we can think up). But that doesn't appear to work, because we already have a bunch of things we want to exclude that would be allowed by such a construction (pocket calculators, ChatGPT, and so on), and we don't consistently require that things inside the boundary have all those characteristics.

So there are certain points along the border that look like they're a "natural boundary" that just reflects some inherent characteristic of the landscape, but we can't generalise that to say that about the whole thing.

Pvt. Parts posted:

Who said anything about human behaviours? I can't tell if you are being overly pedantic or overly human-centric, or both.
We got here because someone posited a "p-zombie AI", and the question about what it would take for people to accept such a thing as intelligent or conscious was raised.

If your point is that fretting about how to draw a box around "intelligence" is a fruitless enterprise, I agree. Specifically because I think it's just an overly human-centric framing.

Negative_Kittens
Apr 8, 2008

[ASK] me about multiple personality disorders
So I was reading about that WoW reddit that tricked a news site that uses AI to produce its articles and it got me thinking about something I was musing back when ChatGPT was in beta: eventually, most of the internet will be rendered impossible to distinguish between AI articles, comments, and content.
When this happens, things like Facebook and Reddit will be basically unusable, since those systems can respond faster and at a greater length than the human users of that website. And then I realized that SA has a pretty decent chance of surviving this, since we have a paywall and active, human content moderation.
Not sure what else to take from this, but I think its funny than a website from 1999 has a better shot of surviving the AI web wars than big giant Silicon Valley companies.

Negative_Kittens fucked around with this message at 23:56 on Jul 21, 2023

Doctor Malaver
May 23, 2007

Ce qui s'est passé t'a rendu plus fort
Why would Facebook become unusable? Most of the time I use it for content from people and organizations I know. Fake content from fake people can reach me only if someone I know shares it or Facebook promotes it.

BrainDance
May 8, 2007

Disco all night long!

Doctor Malaver posted:

Why would Facebook become unusable? Most of the time I use it for content from people and organizations I know. Fake content from fake people can reach me only if someone I know shares it or Facebook promotes it.

I imagine the comment section at least will be ruined.

I dunno I mostly use Facebook because my grandparents and the rest of my extended family do and I moved really far away, so I post stuff about my life for them. But I also use it for some groups in a kind of "since I'm here might as well" way.

The groups are all discussion groups, mostly Linux stuff, and I could see AI ruin them.

Its not the same thing as the bots taking over but when ChatGPT first blew up I noticed a lot of people "answering" people's Linux questions with obviously ripped from ChatGPT answers. Which sucks because it often gives you wrong amswers for that kinda stuff since it doesn't know much after 2021 and software changes, and it makes it sound really confident.

Just a bunch of guys ripping answers from ChatGPT to try to sound smart doesn't ruin a group though, even if it does make it worse. But if it was bots and it was everything ever it could make things pretty bad.

But on the other hand ChatGPT is probably smarter than the average Facebook commenter.

Count Roland
Oct 6, 2013

AI will put more pressure on the algorithm-driven, click-focused models social media uses.

Models which I've always disliked. There is so much hand-wringing about misinformation on the internet these days but I doubt I'll see much of a difference. I've adopted the "on the internet, nobody knows you're a dog" approach since I was young. I go to known sources, known communities for content or information. There is, ultimately, no way to trust what you find online, it all comes down to reputation.

I do hope that things like old school forums will have a bit of a revival in the face of this. Ditto real newspapers, local media. Probably just wishful thinking on my part though. It didn't take chatbots for people to believe nonsense they read online.

Tei
Feb 19, 2011

Negative_Kittens posted:

So I was reading about that WoW reddit that tricked a news site that uses AI to produce its articles and it got me thinking about something I was musing back when ChatGPT was in beta: eventually, most of the internet will be rendered impossible to distinguish between AI articles, comments, and content.
When this happens, things like Facebook and Reddit will be basically unusable, since those systems can respond faster and at a greater length than the human users of that website. And then I realized that SA has a pretty decent chance of surviving this, since we have a paywall and active, human content moderation.
Not sure what else to take from this, but I think its funny than a website from 1999 has a better shot of surviving the AI web wars than big giant Silicon Valley companies.

Email is already in this state. Is used for professional stuff, but seems dead for personal use. Spam and commercial use killed email.
Radio was killed by TV.
Broadband killed cybercafes.
SEO killed Google.

Of course things that are killed this way are killed the masses, these things still exist. Somebody will come here and tell me he still use email to talk with hi niece and a lot of european friends.

Google use to work because information was public, but now is hidden inside Discord and other places. Advertising companies found the key to swamp Google with good looking websites that are shallow tricks to interest people.

If you google "Good headphone", the first pages will be pages with stuff like "Top 10% headphones for games / audiophiles" where 2 of the top articles are advertised this way.

https://www.youtube.com/watch?v=48AOOynnmqU

gurragadon
Jul 28, 2006

Thanks for the video it was interesting. I feel like everyone has noticed that google search kind of sucks now and at least from the video doesn't seem like they are responding to their real problems. At the end when the presenter mentions other ways to search like taking pictures, which is all well and good, but the results are more important. If I take a picture of something it's still going to give me a bunch of junk responses, so I guess the experience was smoother but still useless.

I wish there was a way to get good text results because I'm not a fan of the video style posting that people use on Tik Tok. But clearly younger people are using it because they don't mind video and it's giving them the results they actually want. That may be a replacement for google until advertisers learn how to hijack the platform to get what they want to the front page like they did with google.

AI articles in the vein of the WoW post about the fake story are all over google. My phones google page defaults to giving me recommended articles and stories based on my previous searches. 90% of them are a reddit post with a bunch of AI stuff writing a bunch of fluff that is generically relevant to the topic but not specific in any way to the situation. I don't think google sucking is specifically an AI problem but it sure doesn't help at all.

Edit: The articles almost read like a grade-schooler trying to write a response to a quote. They just have these irrelevant paragraphs that are like intro statements for a book report and are quite jarring when I'm just trying to get information. It's like there trying to pad out word count for a 500 word essay.

gurragadon fucked around with this message at 18:38 on Jul 22, 2023

Owling Howl
Jul 17, 2019

gurragadon posted:

AI articles in the vein of the WoW post about the fake story are all over google. My phones google page defaults to giving me recommended articles and stories based on my previous searches. 90% of them are a reddit post with a bunch of AI stuff writing a bunch of fluff that is generically relevant to the topic but not specific in any way to the situation. I don't think google sucking is specifically an AI problem but it sure doesn't help at all.

It's a general issue across the media landscape. FoldingIdeas on YouTube went in depth on a scam on Audible where grifters look for themes or words high on the bestseller lists and then pay a ghostwriter agency a few thousand to crank out a book about whatever. RedLetterMedia recently touched on something similar with some studios producing huge amounts of super-cheap generic b-movies centered around keywords like "exorcist", "shark", "dinosaur." Words that people associate with a certain genre of movies. I'm sure the same has been happening in music for a while.

So gqrbage books, movies, music, websites and social media posts are being produced in large quantities by grifters ranging from corporate operations to individual basement dwellers using whatever techniques are available to them. It's essentially a more elaborate take on spam email.

Will AI make it worse? I'm sure it will but to me social media is already completely useless. I rarely use Google for anything but locating a gateway to a credible website that can then give me the information I need. I don't buy random books, movies or music before looking for reviews, excerpts or samples. We navigate all these layers of fluff, advertising, scams and grift to get to anything useful and most people do it routinely without thinking about it. It's merely tedious and annoying.

It's people who open emails from Nigerian princes who are ultimately being targetted and it exists because it works.

duodenum
Sep 18, 2005

Tei posted:

information was public, but now is hidden inside Discord

What does this mean? How is information "stored" in Discord? Is it just searching a chat log for a conversation?

Serotoning
Sep 14, 2010

D&D: HASBARA SQUAD
HANG 'EM HIGH


We're fighting human animals and we act accordingly

duodenum posted:

What does this mean? How is information "stored" in Discord? Is it just searching a chat log for a conversation?

Yeah, which you can do, but Discord isn't indexed by big search engines

Rappaport
Oct 2, 2013

duodenum posted:

What does this mean? How is information "stored" in Discord? Is it just searching a chat log for a conversation?

In the before-times, long long ago, people would make websites like Dylan O'Donnell's Nethack spoilers, or wikis, which have now been aggressively taken over by the "fandom" domain, etc., so now if you want to find out the best way to make a pretender god in Dominions you seek out a discord and ask in there. Which is not exactly ideal for niche subjects since Discord is a lovely service in a lot of ways, and openly hostile to searching anything even within servers you're in.

duodenum
Sep 18, 2005

The internet made us stupid, and now we're doing the stupid right back to the internet.

BoldFace
Feb 28, 2011
Many would consider robots destroying social media a net positive for humanity.

Freakazoid_
Jul 5, 2013


Buglord
I haven't really hit any frustrations with google until somewhat recently, when it started not respecting the old tricks I use to refine searches. I've also tried avoiding reddit for the obvious reasons, but it's hard to deny that's where a lot of pertinent information ends up at. I'm glad this forum still does something like that, but obviously not nearly to that extent.

I also hate discord so much because it's designed for privacy, making it easy for a bunch of drama queens to gatekeep you from relevant information until you join in on their bullshit brigade.

Mercury_Storm
Jun 12, 2003

*chomp chomp chomp*
I don't see how information is being lost in Discord any worse than it being lost in MIRC which wasn't indexed either. If you wanted the latest info years ago you would join a chat just like now, except you'd probably get called gamer words for your trouble because lol MIRC nerds.

reignonyourparade
Nov 15, 2012

Mercury_Storm posted:

I don't see how information is being lost in Discord any worse than it being lost in MIRC which wasn't indexed either. If you wanted the latest info years ago you would join a chat just like now, except you'd probably get called gamer words for your trouble because lol MIRC nerds.

the irc people were a level of Online above even the regular Online people by the time i was on the internet and i'm not particularly young, the irc was probably associated with a forum where these things did actually get posted because the forum was busier than the irc itself, and then a few years after that was going into a wiki that at the time would've been fairly well supported by the community

Freakazoid_
Jul 5, 2013


Buglord

Mercury_Storm posted:

I don't see how information is being lost in Discord any worse than it being lost in MIRC which wasn't indexed either. If you wanted the latest info years ago you would join a chat just like now, except you'd probably get called gamer words for your trouble because lol MIRC nerds.

It's all about the timing. IRC was made to fulfill an unmet need in the early years of the internet, is an open communication protocol anyone can use, and shared the height of their popularity with websites, forums and wikis. With some exceptions, there was effectively little information being hoarded on that platform.

Discord is a proprietary business. They need information to be hoarded on their platform to keep people interested. They are doing this at a time where the internet has strongly consolidated to a dozen or so platforms, some of which also do not index their information.

Clarste
Apr 15, 2013

Just how many mistakes have you suffered on the way here?

An uncountable number, to be sure.
To be honest I use Discord constantly and not once have I considered it a repository of information, so I have to wonder where this idea that it relies on that for usage comes from. While there's certainly a kind of institutional knowledge that exists among groups of fans, and maybe people are like pinning big messages or links to google docs or something, if people are regularly using it like that strikes me as a failure of wikis more than a feature of Discord.

The point of Discord is talk to people, preferably your friends or maybe coworkers.

Clarste fucked around with this message at 13:06 on Jul 26, 2023

BrainDance
May 8, 2007

Disco all night long!

irc logs are at least just plain text.

Though, yeah, I am one of those intensely nerdy archive everything types, I've only ever deleted things that I had to because of GDPR and I have chat logs of different things going back to 98 (it's very embarrassing stuff because I was 11.) And I care a lot about preservation. I've been actually going through and organizing my archives, going to burn some of the real old stuff that doesn't exist anywhere else to m-discs, and its been taking months because its that big and that disorganized. It's like 10tb of stuff at this point.

But the first irc channel I ever joined, I didnt have logging on so I don't have any of it left. And that really sucks, and if it had been on a forum or something archive.org would have picked up at least some of it.

But with discord can you even get the logs in any easy way? I googled it real quick and it looks like all the ways are super hacky and require some software outside of discord, and going by how this sort of thing has gone in the past I don't see discords servers as being any kind of long term archival storage.

AOL had those keywords, like websites but not. Like you go to keyword Games and you get a whole thing with message boards and chat rooms and sorta-websites. There was actually a lot of cool stuff that happened there. Nintendo used to do a summer role playing thing in the Zelda universe and the first couple years were on AOL. None of those have been archived outside of a few screenshots. I see Discord going the same way.

Clarste posted:

To be honest I use Discord constantly and not once have I considered it a repository of information, so I have to wonder where this idea that it relies on that for usage comes from. While there's certainly a kind of institutional knowledge that exists among groups of fans, and maybe people are like pinning big messages or something, if people are regularly using it like that strikes me as a failure of wikis more than a feature of Discord.

The point of Discord is talk to people, preferably your friends or maybe coworkers.

Almost all of the ones I'm in are for specific projects, and function like support forums with project news mixed in. It's kinda important for support forums to be on the indexable Internet because other people are going to have the same problems as other people. Discord even has the support for channels that function more as a forum, with topics arranged in a list.

There was one project I did some stuff on, we had an irc channel back like 7 years ago. The irc channel was mostly for the devs to hang out in and talk about stuff, and sometimes people would come in to ask for help but that was mostly on the forum. The project got big, eventually they moved to discord. Now, most support is done on the discord and the forums are pretty dead. If those threads were on the normal Internet where google could pick them up it would probably help a lot of people.

I dont think the point of Discord is just to talk to people like friends and coworkers since they've very specifically built in a lot of the functionality of forums that isn't really aimed at that.

Like this, these are the support forums for sd.next


which sucks

BrainDance fucked around with this message at 13:16 on Jul 26, 2023

Gynovore
Jun 17, 2009

Forget your RoboCoX or your StickyCoX or your EvilCoX, MY CoX has Blinking Bewbs!

WHY IS THIS GAME DEAD?!
Bear in mind, Discord started out as a low-footprint way for gamers to shout "LOLOL!" and "Pwnd!" at each other. Serving as a large bulletin board is kinda off-label.

Clarste
Apr 15, 2013

Just how many mistakes have you suffered on the way here?

An uncountable number, to be sure.

Gynovore posted:

Bear in mind, Discord started out as a low-footprint way for gamers to shout "LOLOL!" and "Pwnd!" at each other. Serving as a large bulletin board is kinda off-label.

More specifically, it was for MMO guilds to use as an out-of-game guild chat with a built-in voice chat for raiding.

KillHour
Oct 28, 2007


I was legit like "who the hell is using Discord as a source of information" and then I remembered that I ran into issues setting up Moonlight streaming to my Shield and I had to join their Discord to figure out the issue. On the other hand, they were very fast to respond and a PR fixing the problem was committed less than a day later instead of sitting in a GitHub issues page for a month.

It's me, I'm part of the problem.

Edit: Weirdly enough, I really like how Teams threads chats inside a channel. Having a system to automatically archive those in a publicly searchable way would probably be the ideal combination for near-real-time discussion that needs to be archived. Someone should build that.

KillHour fucked around with this message at 14:13 on Jul 26, 2023

SCheeseman
Apr 23, 2003

With Disney and the media industry, the entities responsible for the current copyright and IP regulatory frameworks over the last 100 years going all-in on generative AI, the idea that copyright will in any way be helpful in serving the rights of affected artists and workers is an absolute joke.

Tei
Feb 19, 2011

Mercury_Storm posted:

I don't see how information is being lost in Discord any worse than it being lost in MIRC which wasn't indexed either. If you wanted the latest info years ago you would join a chat just like now, except you'd probably get called gamer words for your trouble because lol MIRC nerds.

Because Discord is not a pure chat system. People wrote long explanations and guides that take effort. Sometimes you ask something, and it is asked after 9 hours.

The "format" of discord is somewhere between IRC and a PHPBB style forum like this one.


The comparison fails flat, because when IRC was huge, also where forums, and forums would be indexed. Nobody would have wrote you a long guide in IRC, instead they would reserve that kind of effort for a post in a forum.

We don't have forums anymore because people has been leeched by social networks with the lure of "bigger city, shiny lights, more eyeballs for your post, maybe meet new people".


Now, this will affect negativelly systems like ChatGPT, because they will have less content to leech. Except if somehow ChatGPT is allowed to read backups of Discord logs. I don't know if that is possible.

Mederlock
Jun 23, 2012

You won't recognize Canada when I'm through with it
Grimey Drawer
Oh, you just know that they'll sell that data at the drop of a hat to the right buyer at the right price

Tree Reformat
Apr 2, 2022

by Fluffdaddy
There absolutely was a kerfuffle about Discord changing their ToS months back to make data scraping (possibly for ML training) easier, and more recently introducing an AI helper agent into servers.

MixMasterMalaria
Jul 26, 2007

Mederlock posted:

Oh, you just know that they'll sell that data at the drop of a hat to the right buyer at the right price

https://www.youtube.com/watch?v=vWRlxSGf_ns

Adbot
ADBOT LOVES YOU

khwarezm
Oct 26, 2010

Deal with it.
Seeing all this AI art proliferation has got to be one of the most depressing things I've seen in my life, especially since so much of my life and social circle is centred around art and artists. Apologies for being such a drat doomer but the thought that the future is just going to be art itself, the most personal and human thing you can have, just turn into the product of unthinking algorithms spitting out facsimiles of the work of actual humans mashed from thousands of images on the internet while some Silicon Valley fucker creams himself over this being the future of humanity makes me genuinely want to live in a cave in the woods.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply