should one accept A.I in some form right now?

[OppositeKeith]

Recently with all of the rapid Advancements that A.I has been getting that has been starting to effect Entertaiment careers like writing, music, art, voice acting, etc. i slowly go into a downward spiral on what should i do right now that this sort of thing is now possible and viably accessible to anybody with a computer at their home.

Questions like:
Should i continue making and improving art?

will their be a career for this for me when i finally get good enough to start one?

even if a solution is made that both sides of argument agree with, will these industries still die out because of the accessibility?

should i even use A.I services that are very charitable and secure?

I don't really keep up with the news, i have my own life i need to worry about most of the time. i don't actively engage in these arguments as its hard to swallow and deal with people who think this is really great for everybody especially when it goes off the rails and treat me like im an idiot, i want to keep my nose clean.

my current opinion about it so far is something like this: i should try and continue to do it my art and music cause it is something that makes me happy and have a sense of purpose to do, A.I, as flawed as it is right now, is now here and will improve in the coming years by people who dont see the dedication and hard work that goes into careers outside of their own and i should realize and accept that. i wish the best of luck to those against A.I and those who are going forward with it to find some agreement with people like me although admittedly i doubt it.

maybe i have flaws in my opinion/philosophy that i dont see yet, hell i may as well have answered my own question in a way, but i'm curious to hear this being discussed here.

[madness]

those factory workers who rebelled and destroyed machines decades ago did so because because they felt their mode of survival was being threatened by technological innovation. it is understandable why at a surface level the easiest target is the technological advancement itself, far harder to blame the underlying structures which need to be changed in order for us to correctly utilize new technologies.

it seems obvious to me that personal skills and hobbies will always have a place, but the anxiety only really starts when it involves earning a living. I can always find a reason for practicing creativity outside of a marketplace, can you? I would love for AI to do all the work for me, to save me hours on something, because then I can go work on something that AI cannot do, forever and always.

so what's the problem? I do think it is income insecurity at its core, just like it has always been with "automation". but maybe I am missing something else. human creativity is boundless with an unfettered mind. the endless debates about the nature of AI go nowhere, ask instead why life isn't as great as our current technology has the potential to make it.

[OppositeKeith]

it seems obvious to me that personal skills and hobbies will always have a place, but the anxiety only really starts when it involves earning a living. I can always find a reason for practicing creativity outside of a marketplace, can you? I would love for AI to do all the work for me, to save me hours on something, because then I can go work on something that AI cannot do, forever and always.

I'm sure it will have a place for many, possibly similar to things like music where despite the advancements of technology around it and equipment people still have enjoy the feeling of playing an instrument or seeing someone or some group playing a song with instruments too. earning a living or getting some form of living at all feels anxious more than ever and half of the time i feel like i have no know choice but to use the best of my skills to make meets end, that being said i can make reasons to practice them outside of the marketplace. i like to practice more cause i enjoy the experience and i make that experience possible by doing it with ideas that flourish in my head, ideas that i get from others who aspire to try it to begin it. A.I may do most of the work for me but half of me wants me to do it myself cause there is aspects of life that i feel like i give up too easily on and don't go into it with an open mind. makes me think all too many times.

[NULLinvis]

my current opinion about it so far is something like this: i should try and continue to do it my art and music cause it is something that makes me happy and have a sense of purpose to do, A.I, as flawed as it is right now, is now here and will improve in the coming years by people who dont see the dedication and hard work that goes into careers outside of their own and i should realize and accept that. i wish the best of luck to those against A.I and those who are going forward with it to find some agreement with people like me although admittedly i doubt it.

For sure, art should be created to express ones self and to further understand ones role in this chaotic and uncaring world this is something a machine will never be able to replicate.

in practical terms as well, I like to think AI will never replace human artists. AI must leverage a massive labelled training set of data to be able to generate good outputs. If we are using DALLE as an example, I can type in "X in the style of Pollock" and it will generate a poor knock off in his style, due to the vast amount of categorised and labelled data online pertaining to Pollock and his work. If you try this with someone more niche or contemporary, the low volume / lack of data in the training set will just mean you get garbage out. This is encouraging, as even though an AI could produce a painting in the style of Pollock it will never create a Pollock - there are no means for it to process and interpret a unique view on the canon of Art that a living artists could and emotionally synthesize into a work. A living artist also has to understand their material conditions (consciously or not) in such a way that their art interfaces with the cultural zeitgeist in a meaningful and boundary pushing way that resonates with others, something an AI would never be able to do.

[Blog47177]

Ai is hyped for now until we need to find out more that our technology does not have the skills to do for now.

I hope we can tame AI at some point.

[GaryStu]

Robots automated the generic animu of the Deviantart front-page. big lol, imo.

To me, the biggest threat of AI art is not the art itself, but the lengths people are willing to go to stop it. Pro-copyright hawks are circling around AI art, and they're who worries me. Any expansion to copyright law will hurt art a lot more than AI could ever. So I've found myself in the position of defending AI art, even when tbh I don't really like the idea of using it as a technology.

Like its just kind of boring to use. If I have a specific image in mind its still quicker and easier for me to just draw it. And I'm still saying that when a digital illustration can take me eight hours lol.

When I have a better graphics card, and as the tech improves, I am interested in how I can intergrate AI into my workflow. But now, it's grating to include in my workflow so I don't.

- Should i continue making and improving art?
You should keep making art.

-.will their be a career for this for me when i finally get good enough to start one?
AI art can be used as part of industry art workflow. You can use it as photobashing material, and guide it to an extreme extent. There's a strong possibility it becomes an industry standard, but having strong underlying art fundamentals will still be key.

- even if a solution is made that both sides of argument agree with, will these industries still die out because of the accessibility?
Why would something die out because of accessibility?
I don't want to come to a compromise with copyright law supporters. Any compromise is a loss.

- should i even use A.I services that are very charitable and secure?
I use Artbot, which relies on the distributed computing of the Stable Horde. It is free but incredibly insecure. Its transparent about its insecurities though, and I'm comfortable with them for my purposes.
I only use Stable Diffusion on principle. I like FOSS, and think its important to use FOSS over, for example, an Elon Musk funded project.
I think anyone interested in trying AI art should look into Stable Diffusion first. If you have a graphics card that's good, you can run it locally and offline. It doesn't get more secure than that!

So to end off...
This isn't a value judgement on AI art itself. A lot of it sucks ass. This is how I feel about the tech, the art produced by AI itself by-and-large makes me feel... well, the same way I feel looking at generic animu lol. i feel fucking nothing.

[nightwolf334]

Let's take music as an example. People have always complained that popular music is all samey. This has been happening since the extreme commercialisation of music in the 60s. Now, how much of what is commercially successful music actually makes up of how much music is out there? Less than 1%. There's a band out there that would be your favourite, and there's a good chance you will never discover them. That's how deep the sea is. AI, in my view will not change this dynamic. Yeah, there will be AI music that will come out and threaten the top commercial-echelons, but consider those that already make music that don't do that. You could probably extrapolate this to all the arts, and I'm pretty confident its the same.

[Blog47177]

https://www.thecrimson.com/article/2023 ... n-webinar/


Well Sal Khan of Khan Academy has some good points on AI and how we should look into this.

[yequari]

---

[Blog47177]

So far AI is going through the Overhype phase where ChatGPT because Venture Capitalists want investors to put money on AI. However to really understand its impact in practice we really need to wait for a dot-com bust type event or a crypto investments type bust to really find out how we can adapt to AI tech or web 3.0.

[sixeyes]

Another discussion about these things, that was kind of drowned out by these shiny image/text-generators, is around the suddenly widespread ability to do blackbox pattern matching. Huge datasets can be condensed into relatively small programs that'll classify your stuff for you, without anyone knowing how it actually does that.

So, you get near-instant detection of whatever you train the thing on, sometimes from what's "noise" to you or another human. Used already in medecine (iirc, for diagnosing skin cancer). But the way it comes to that conclusion is hidden, you'll just have to trust it. And like, these models are expensive to make and require decent hardware to run after the fact. Who has access to that? What models will actually be produced, if capital is what sets the course? How and where will they be deployed?

This is sci-fi but i'm worried about things like "lie detectors" and such.

[Blog47177]

Update Google is hyping up their answer to Chat GPT. Note we need to wait and see how this plays out.

[windigo]

Recently with all of the rapid Advancements that A.I has been getting that has been starting to effect Entertaiment careers like writing, music, art, voice acting, etc. i slowly go into a downward spiral on what should i do right now that this sort of thing is now possible and viably accessible to anybody with a computer at their home.

Questions like:
Should i continue making and improving art?

will their be a career for this for me when i finally get good enough to start one?

even if a solution is made that both sides of argument agree with, will these industries still die out because of the accessibility?

should i even use A.I services that are very charitable and secure?

I don't really keep up with the news, i have my own life i need to worry about most of the time. i don't actively engage in these arguments as its hard to swallow and deal with people who think this is really great for everybody especially when it goes off the rails and treat me like im an idiot, i want to keep my nose clean.

my current opinion about it so far is something like this: i should try and continue to do it my art and music cause it is something that makes me happy and have a sense of purpose to do, A.I, as flawed as it is right now, is now here and will improve in the coming years by people who dont see the dedication and hard work that goes into careers outside of their own and i should realize and accept that. i wish the best of luck to those against A.I and those who are going forward with it to find some agreement with people like me although admittedly i doubt it.

maybe i have flaws in my opinion/philosophy that i dont see yet, hell i may as well have answered my own question in a way, but i'm curious to hear this being discussed here.

I think AI should mostly be used for fun things (like replicating a voice actor's voice for a meme or overall low quality content) instead of being used for other things.
This video is a great example
I tested out Chat GPT myself and I feel as if it's more or less a lobotomised version of what it could truly become, people shouldn't be relying on such an AI since it's clearly made to only answer on certain beliefs. (it's a rumour but some people seem to be getting the wrong answer for technical things)
At this stage I'm not sure how a fully fledged AI would act but the AIs which are the disposal of the public aren't the 'job replacers' that many people fear them to be. It can do great if it's trained at a certain task but it certainly can't be flexible on its own, human input is still needed to train the AI and to moderate it.
In the end, AI could be great certain automatic jobs but it certainly isn't ready for anything too complex. (like straight out replacing certain artists, from what I've seen that AIs can't depict small details in their art and they keep messing up random things like an eye colour, so at best it can only replace the mediocre artists or be used as a tool to enhance an artist's view)

[King Barnaby]

I got access to DALL-E in July (had to wait 3 months on the waitlist), I made some fun dog paintings you can see at https://imgur.com/gallery/D5Z9gL8. It didn't stop me from creating my own work, in fact it greatly inspired me, and I could also make certain prompts for inspiration, and I sometimes every now and then do the same for my own work which I do manually.

As for ChatGPT, it's alright but yes I can personally attest that it is horrible in technical details, for example, you can make it say wrong things about math if you hide it, you wouldn't get results with "Are 8 and 6 prime numbers" but I managed to get it to say exactly that when I asked "prove that all primes are the sum of 3 cubes". It also got certain details wrong when I asked to paraphrase, etc. Also anyone with experience can tell whether a text is made by ChatGPT, it feels soulless.

[Blog47177]

https://apnews.com/article/artificial-i ... 5119a69962


Apparently Elon Musk decided to insert himself in the AI Debates. First of all we need to wait for congress to insert themselves in the AI Debates after they finish the Crypto and Tik Tok debates.

Are tech companies moving too fast in rolling out powerful artificial intelligence technology that could one day outsmart humans?

That’s the conclusion of a group of prominent computer scientists and other tech industry notables such as Elon Musk and Apple co-founder Steve Wozniak who are calling for a 6-month pause to consider the risks.

Their petition published Wednesday is a response to San Francisco startup OpenAI’s recent release of GPT-4, a more advanced successor to its widely-used AI chatbot ChatGPT that helped spark a race among tech giants Microsoft and Google to unveil similar applications.
The letter warns that AI systems with “human-competitive intelligence can pose profound risks to society and humanity” — from flooding the internet with disinformation and automating away jobs to more catastrophic future risks out of the realms of science fiction.

It says “recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control.”

https://www.reuters.com/technology/ital ... 023-03-31/


Here is more Italy wants to investigate ChatGPT over data collection.

[Laz1985]

There's a problem here, and that's the assumption that so-called AI artists are artificial intelligence in the first place. A.I. learns. It develops an understanding. Most "AI art" is created by a simple algorithm that searches for the keywords you type in, then samples the most common input for each pixel. This is why you have weird numbers of fingers and occasionally outright body horror in most "AI art." Furthermore, because artists did not give permission for these algorithms to use their art, the issues of consent and copyright come into play.

In short, once actually artifically-intelligent art programs are here, I'll be down with it. But for now, a simple algorithm that can't learn is being used to essentially steal artists' work and rebrand it. That's not cool.

[Rynn]

For sure, art should be created to express ones self and to further understand ones role in this chaotic and uncaring world this is something a machine will never be able to replicate.

in practical terms as well, I like to think AI will never replace human artists. AI must leverage a massive labelled training set of data to be able to generate good outputs. If we are using DALLE as an example, I can type in "X in the style of Pollock" and it will generate a poor knock off in his style, due to the vast amount of categorised and labelled data online pertaining to Pollock and his work. If you try this with someone more niche or contemporary, the low volume / lack of data in the training set will just mean you get garbage out. This is encouraging, as even though an AI could produce a painting in the style of Pollock it will never create a Pollock - there are no means for it to process and interpret a unique view on the canon of Art that a living artists could and emotionally synthesize into a work. A living artist also has to understand their material conditions (consciously or not) in such a way that their art interfaces with the cultural zeitgeist in a meaningful and boundary pushing way that resonates with others, something an AI would never be able to do.

With the current advancements in LLMs and other models, AI art will be indistinguishable from human art within the next decade easily.

This is encouraging, as even though an AI could produce a painting in the style of Pollock it will never create a Pollock - there are no means for it to process and interpret a unique view on the canon of Art that a living artists could and emotionally synthesize into a work

Human's don't do this either; at a basic level humans are simply an advanced algorithm that is the result of their biology, history, and environment, all of which can apply to machines just as easily as they do people. Biology = hardware, history + environment = software.

A living artist also has to understand their material conditions (consciously or not) in such a way that their art interfaces with the cultural zeitgeist in a meaningful and boundary pushing way that resonates with others, something an AI would never be able to do.

This misses that the above is a result of the way people view the art, which may or may not line up with the intention of the artist. I look forward to some games of "Human or computer-generated art?" in the future.

[vincent]

[snipped rest of post]

Human's don't do this either; at a basic level humans are simply an advanced algorithm that is the result of their biology, history, and environment, all of which can apply to machines just as easily as they do people. Biology = hardware, history + environment = software.

I think this is a really reductive way to look at humanity.

Can a machine be inspired by something? At current moment, no. A machine can take a bunch of images, take common features, and re-create them in specific patterns. A machine cannot be truly inspired by a work of art. A machine cannot see a Van Gogh or a Pollock as anything but a collection of brushstrokes and paint spatters that it can recreate.

A human, however, can see those works of art as something beyond re-creating brush strokes. We can look at those works of art and be inspired by them to create something wholly original from those works. It is the same way that tracing an image is different than looking at an image and capturing the heart of it and using that within your work.

I also don't think that a machine can have "history" and "environment" in the same way a human does. A machine has no backstory or life with which to draw inspiration from to create art. To view the human mind as entirely algorithmic would suggest that two people with the same history and environment and biology (siblings, perhaps?) would inherently turn out the same: clearly they don't. There is something further going on there within the human mind that an AI cannot recreate.

Emotional synthesis is something so beyond just an algorithm of "this happens = I feel this way = I make this art because I feel this way" that I don't see any way that a machine can ever replicate that in a way that matters.

I think we'll see a lot of corporate art made by AI in the future, maybe. I don't see AI art having a fundamental effect on the artistic landscape in the same way that like, photography changed art.

As a final note, I just want to bring up something anecdotally from my studies (I'm an Art History major and about to graduate with my Bachelor's):

"The arts could save themselves from this leveling down only by demonstrating that the kind of experience they provided was valuable in its own right and not to be obtained from any other kind of activity.

Each art, it turned out, had to perform this demonstration on its own account. What had to be exhibited was not only that which was unique and irreducible in art in general, but also that which was unique and irreducible in each particular art. Each art had to determine, through its own operations and works, the effects exclusive to itself. By doing so it would, to be sure, narrow its area of competence, but at the same time it would make its possession of that area all the more certain"

This is a quote from Clement Greenberg's Modernist Painting written in the 1960s. Within it, Greenberg discusses how Modernist painting was able to set itself apart due to the ways that it establishes what painting can do that no other mediums can do.

I see this happening to AI art within the near future. At some point, we are going to see AI art forced to contend with what it can do that other forms of art cannot do. At the moment, that is very very little. I doubt this will change much in the future. While Greenberg's account of modernist painting is hardly the only one, I think this quote is something to hold onto when discussing AI art: AI art is going to, one day, be asked to stand on its own. Will it hold up? Only time will tell.

[Blog47177]

https://www.reuters.com/technology/bill ... 023-04-04/

Bill Gates now responds to the AI debates.

[Blog47177]

https://www.reuters.com/technology/aust ... 023-04-05/


Well a certain politician in Australia is accusing Chat GPT for Defamation.

[NULLinvis]

Human's don't do this either; at a basic level humans are simply an advanced algorithm that is the result of their biology, history, and environment, all of which can apply to machines just as easily as they do people. Biology = hardware, history + environment = software.

You completely misunderstand / misconstrue my point, and respond with incredibly inane tech bro hur dur human brain just like computer.

AI is just a probabilistic, specialised network model trained on a specific set of labelled data. This can be text, images, etc. During training this data is continuously passed through the network, and the output is rated on how 'correct' it is (this is called supervised learning and compsci students will know the pain of using MATLAB and Fisher's Iris dataset to classify flowers). The AI takes this output and the rating of the output and passes it back through the network, making adjustment to its parameters against costing functions such that the new output is closer to the correct answer.

Because of this, AI is great at classifying data with a strong underlying probabilistic function, or generating new data points based on its corpus of training data such as Chat-GPT responding to text prompts or DALL-E making images from text prompts.

As such it's performance is inherently chained to this availability of high quality data, which is what i was getting at in my previous post. Yes I can tell DALL-E to create an imitation pollock, but it will never create a pollock insofar that it won't be able to generate a piece that is so of it's moment, a cumulus of that time's social zietgiest, that it would be in essence a pollock - a radical, new-art that will become a permanent moment in the canon of contemporary art. There is also a reason it has its own 'style' (i.e. shitty digital DeviantArt art), and that is because the mass of its training data, presumably scraped from the internet, features a lot of well labelled, shitty digital DeviantArt art. It doesn't think, unconsciously or consciously, like a human does, it just segments a prompt into tokens and transforms them into an n-dimensional feature-space to be passed through a network with pre-trained parameters.

And spare me "this is exactly what humans do tho" shit, because if you are being reductive then yes it is, but that doesn't take into account anything that makes us human - our ephemeral sense of the 'I' through which all information form the world is filtered, formed early in our lives after the Mirror Stage and continuously shaped through the signs we learn and the experiences we have across our life, and how this 'I' effects all the outputs we produced (even this is a reductive statement, but it is a lot better than brain is computer brr). An AI will never have this, as it is neither thinking or self aware.

Stop falling for the AI hype / scare train, (a hauntology of the metaverse perhaps?) Goodbye.

[NULLinvis]

"The arts could save themselves from this leveling down only by demonstrating that the kind of experience they provided was valuable in its own right and not to be obtained from any other kind of activity.

Each art, it turned out, had to perform this demonstration on its own account. What had to be exhibited was not only that which was unique and irreducible in art in general, but also that which was unique and irreducible in each particular art. Each art had to determine, through its own operations and works, the effects exclusive to itself. By doing so it would, to be sure, narrow its area of competence, but at the same time it would make its possession of that area all the more certain"

This is a quote from Clement Greenberg's Modernist Painting written in the 1960s. Within it, Greenberg discusses how Modernist painting was able to set itself apart due to the ways that it establishes what painting can do that no other mediums can do.

I see this happening to AI art within the near future. At some point, we are going to see AI art forced to contend with what it can do that other forms of art cannot do. At the moment, that is very very little. I doubt this will change much in the future. While Greenberg's account of modernist painting is hardly the only one, I think this quote is something to hold onto when discussing AI art: AI art is going to, one day, be asked to stand on its own. Will it hold up? Only time will tell.

Amazing quote and great analysis

[Blog47177]

Here is a segment on 60 Minutes about AI as the debate rages on.