Can artificial intelligence art ever really replicate the human artist's emotional authenticity?

Can artificial intelligence art ever really replicate the human artist's emotional authenticity?

  • Yes

    Votes: 2 25.0%
  • No

    Votes: 4 50.0%
  • Unsure

    Votes: 2 25.0%

  • Total voters
    8
Thank you for taking the time to respond to me and I'm glad you jumped in as you make a really good point,

"Any Art makes you emphasize, it is with the art piece not necessarily the author."

This is where for me it gets very interesting, if AI is able to produce the emotional rawness similar to say, Tracey Emins narratives and the viewers emotional reaction doesn't need or require the author to understand the work, would the knowledge of the complexities of human realness to the experienced by the author make the art more emotionally authentic to the viewer?

You have got me pondering on my own question, Thank you!

H.
I need more pondering time for this. For me, Nick Cave made some great points. So, do we just make a new category of art. And let works stand up and be counted for what they are.
 
I think the question needs tweaking...
When art-ificial intelligence art etc etc

My tuppenceworth...When our AI genuinely experiences that moment of fear, when leaning back on two legs of an office chair and about to fall...it will be a real turning point.
 
Art has many aspects. Some of them are purely technical and communicative, for example: Fulfilling certain aesthetic standards ("being nice to look at"), or conveying a mood or message (a painting makes the viewer sad because it depicts a mother mourning her child). In these aspects, AI can already "replace" human artists or will be able to in the near future (something like "throwaway illustrations" for magazines or web media merely exist to catch the reader's eye and have some basic symbolic meaning).

However, "fine art" (i.e. art that's disconnected from practical concerns and purely exists for itself) has human self-revelation, human self-realisation and human communication at a deeper level than relaying a simple message at its center. The human aspect is what gives this type of art its meaning. Even now, there exist thousands of AI-generated "artworks" that are probably wonderful to look at, but I'm not really interested in looking at them. Why? Because to me, art isn't just about pretty pictures or technical skill. It's also about the human aspect, and while AI continues to evolve, the importance of that aspect will only get bigger.

Do I deeply think about the artists emotions and experiences of every painting that I look at? No. But the fact that there's a human on the other end of it is still essential. If a painting resonates with me, it could just be because the brushstrokes are perfect or the symbolism is so striking. But it could also be because I feel something that's connecting me and the artist, who has gone out of their way and spent time and effort to create something with their own hands. When the artwork is the result of an algorithm deciding on what kind of patterns are most commonly associated with some word prompt, that feeling simply isn't there.

AI will only be able to "replicate" this quality of art when we're talking about something like a robot can basically be considered a human being.

Your post reminded me of something. About a month ago, I did an image search of Kay Nielsen's artwork because it's just so beautiful and I sometimes need to look at it. When I did, I noticed a couple "off" images in the image search I hadn't seen before. At a very quick glance, they looked just like Kay Nielsen's work, but when I clicked on them and looked at them more closely, I saw they were flatter, more modern looking, and just lacked that certain "something" that can be seen in his other works. I then noticed they were actually AI images done in his style, which led to my discovering a whole collection of AI generated images in his style. While the AI generated images were pretty cool to look at, they also read like knock offs to me. They definitely don't capture whatever it is - that hard to define thing - that makes an artwork sincerely Kay Nielsen's. Part of this could be due to Kay Nielsen having worked in traditional mediums, maybe that makes it harder to replicate via AI, but I'm not sure.
 
No.
I am over on DeviantArt and use the DreamUp AI since discovering it after returning to DA after 9 years. Whatever prompts I use the paintings are not.....so my answer is no.
I have some of my own sketchbook pages on my gallery there also.
Username BelleBeryl same paints avatar as here.
 
No.
I am over on DeviantArt and use the DreamUp AI since discovering it after returning to DA after 9 years. Whatever prompts I use the paintings are not.....so my answer is no.
I have some of my own sketchbook pages on my gallery there also.
Username BelleBeryl same paints avatar as here.
Thank you for taking the time to respond, I look forward to seeing your sketchbooks on your gallery.

H.
 
something like "throwaway illustrations" for magazines or web media merely exist to catch the reader's eye and have some basic symbolic meaning
Here you touch on an interesting previous topic.

Just witness journals. Now journalists take a cell phone to do everything and they are expected to make do with it. Publishers do not really care about the quality of the illustration as long as it is an illustration. In the run for profit they take amateur pictures if available to save on costs. And as long as there are readers they won't care.

The thing is, that is utilitarian, throw-away (like a normal stone in the stone age, good to throw away, then and now), but it is not the same as art or craft (neither is a stone the same as a lithic utensil). The fact that one can do away throwing a stone, or a quick snapshot, a scribble or a screech, does not imply that haphazard works are art.

Art, like a stone knife or arrow, stems from a difficult process of internalization or comprehension. Yes, so did throwing the first stone. Not, seeing someone else do it and repeating it, does not. It is this process that is difficult. And the more difficult it is, the more we appreciate it, even if later reproducing it is trivial.

The difficulty stems from that process, where one collects information, processes it, re-works it and produces something. Much like an AI. But in difference, when we do (and when we appreciate the result), we apply specific criteria, because those of us who didn't, perished, and we are the result of a billion years of selection on changing criteria, fighting for survival, interacting with Reality, and that is also encoded in our genes and brains, and on top, we are constantly suffering mutations, changing, as is the environment, our life always at stake. And it is not one ChatGPT, it is ten billion human ChatGPTs only now (ignoring history) and competing. That continuous, billion year process has refined in all living things patterns for survival, patterns that we, who communicate, strive to grasp and express because they are important for us. And we have evolved to like what over that billion years has proven useful and advantageous and dislike what didn't. We appreciate/like "expressions" that appeal deep in our "soul", that we "feel" are "right", we recognize it when we see it, even if we cannot say what it is that defines it, in how fishes move in a fish bowl, birds fly or flee, or even if a car moves with a sense of urge, and so does a lion or dog feel our fear or that of its prey; and when we do art, we express a pattern deeply and subconsciously ingrained in our deep self, so deep we weren't aware it did exist.

So, AI.

Now AI is like a monkey that has seen other throw a stone: it is easy to repeat the gesture, but it hasn't gone through the creation agony as the first monkey did. Mondrian was a genius. Kandinsky. So many. But if I go now, paint a canvas green and claim it a work of art just because I saw their works, I am actually not creating. Nor is the AI. If I fight to grasp that inner drive, that I cannot express, and strive to find the words, music, painting, etc to make it raise to conscience, mine and others', then I am creating, and if I can do it in a way others understand, better. Etc.

It does not mean theoretically AI cannot. But it would need a reason to live, and a long selection to survive and develop that strength and interest and skills by itself, many by trial and error, and then it would need a changing environment, to retrain, throw away, make mistakes... then it would have something to express. Only, most likely, it being an AI, it would have something we cannot understand for its requirements and skills, and interests, and likes/dislikes would be different. Without a shared evolution and interests, we wouldn't be able to appreciate AI art (hint for AI programmers: just do anything and claim we don't understand it because we are humans, and get rich selling the emperor's clothes, hint for the rest, don't believe them).

Maybe one day, AI will reach the singularity and be self-conscious. That is not enough. It will have to evolve and develop that urge to live, and find its own strategies. Then, maybe it will become robots. And then maybe we become colleagues working hand in hand. And then one day, we will have a shared experience, distilled by a long selective process, which led us to similar likes and dislikes. Maybe that day, AI will make Art that humans can appreciate.

So, yes, one day, maybe, theoretically, AI can be a creative artist we can appreciate. But that is long away.

Meanwhile, can it just mimic what it sees off the Internet? Of course it can. My printer/scanner also can. No mystery, And if I work my printer to run off ink, I may get interesting copies by chance that are different from the original, but appealing as well. Just like a flute sounded by animals by pure chance.

And note: I already mentioned two child books. These questions have been pondered for far long than modern AI. You can trace back them to children tale collections from three thousand years ago. It is also the eternal question of what makes Art and whether a scammer can fool someone.

Of course, we are not so much different from an AI, other than a billion years of shared evolution. And of course there will be ambitious scammers taking shortcuts to sell us that any John/Jane Doe with a cell phone can make pictures like Cappa, or an AI music and Art like Leonardo or Bach. If you want to believe that, that is your choice, and BTW I have a bridge in Brooklyn for sale. And note as well, scams work every day on lots of people. That doesn't make them any more legit.
 
Last edited:
I am a trainer and ... my delegates tell me they learn more and prefer my training because I don't read every single slide...
I bring my personality to the classroom, I will deviate from topic, do something unexpected like..play harmonica for a second or ask a random question like 'what word rhymes with orange'?

Initially AI art will likely create exceptional works to high standards over and over.

When AI learns feel, and can hear what is not being said, then it will truly ascend into a new realm because Feel and sense is a human prowess, AI will then realise and want to elevate itself above human senses.
 
While the AI generated images were pretty cool to look at, they also read like knock offs to me. They definitely don't capture whatever it is - that hard to define thing - that makes an artwork sincerely Kay Nielsen's. Part of this could be due to Kay Nielsen having worked in traditional mediums, maybe that makes it harder to replicate via AI, but I'm not sure.
Yeah, models that haven't been trained on the particularities of any given artist's style will usually have trace elements of that somewhat airbrushed semi-realistic digital illustration style that AI image gen defaults to, and that can be a tell. But if the person creating the model really wants to, they can do stuff like make it look like the image has real brushstrokes etc. I've "accidentally" saved AI paintings on Pinterest multiple times, so AI is definitely already good enough to "trick" people, especially at a glance.

But that's the thing - the only time when I'm in any way impressed by AI images is if they "trick" me. If I know that's AI right from the start, or I later find out a painting that I looked at was AI, its value to me drops steeply, no matter how perfect it was. Apparently many "AI artists" are aware of that fact, because many of them go out of their way to hide (or at least not mention) the fact that they're using AI. I think this points at a deeper layer of what art means to people. Sure, good technique, interesting concepts etc. are important. But another big thing is meaning, and I'd say that meaning is grounded in a shared human experience. And I would say that using your body, time, emotions etc. to create art directly (or maybe as directly as possible, because at the end of your day your own mind and the tools you use are still a filter), the feeling of that shared experience is much stronger.

Some might say AI is just another tool (like a brush or a pencil) that people can use to express that meaning, and they might have a point ... but to me, the difference between human-made art and AI images is like the difference between a photo that was passed through an Instagram filter and real, physical makeup that looks identical. The former looks nice, but the connection to the person is a lot more shallow. They just put on the filter because it was easy. If someone goes out of their way and puts physical and emotional "labour" into creating the perfect makeup, there's a sense of something deeper. Yeah, maybe they just put it on because they wanted to look nice. But what made them go that far? What exactly was it that made them want to invest all the time and effort, what drove them? this additional "human" layer adds something that I haven't seen in AI images so far.

Because of this, I'm not the biggest fan of the "looking for mistakes and pointing them out" approach to AI image criticism. AI will inevitably get better, and humans make mistakes too. For me it's all about the human connection, even though it's often in the background.
 
Hi my name is Henry, new to Creative Spark, but finding it a really inspiring place to visit. I am currently researching art and emotional authenticity for a Fine Art thesis, asking; Can artificial intelligence ever really replace an artist’s human emotional authenticity?

I am finding it a really interesting topic, filled with passionate debate from all sides and I am gaining some extensive and thought provoking knowledge through my interactions with all types of incredible artists, from those who choose to work with more traditional methods to those embracing newer technologies.

This poll has just 3 responses, Yes, No, Unsure, but please feel free to add or send me any additional thoughts as all comments will be greatly valued.

Thank you to everyone who supports me and takes part.

Kind Regards

H.
I am new to creative spark as well and I find your question really interesting. I think there may be a place for AI, where that fits might be different for everybody. When I think of this question it takes me for some reason to cooking because everyone says you can taste the love that’s put into a dish, and I think that is so so true.It may be that some pre- packaged spice mix or sauce may be involved, but if you enjoyed the process of cooking and your intention is to serve a beautiful meal then that becomes part of the experience of those enjoying your dish.I kind of think whatever your creating if your putting yourself into that wholeheartedly using whatever tools you have then that may be felt by people who connect with the work.
 
Last edited:
Performance has a point. Can't remember the movie, one were a teacher told a girl who played perfectly she had no future in music and a terrible performer she had the potential to become the greatest performer of all time. The first was mechanically perfect, the second was temperamental. That makes the point: you can feed as much as you want to a machine, it will interpret the score or a made up picture perfectly.

Now, I've pondered this for decades: can I simply make the machine make small deviations so it seems like done by a human? The problem is humans do not make the mistakes at random, it is temperamental. We all know when a line is uncertain, straight, has flair or is inspired, we can see a line and call it sad or happy. But we do not know why it does. It is not just random deviations, or we woudn't be able to tell, there is a pattern to them. An artist can feel the pattern and rework it, maybe not knowing what it was they perceived, but most often intentionally. So, if I, or programmers want an AI that does the same as a human, one should know first what is it that makes us notice these differences, which we do not know, and hence cannot teach. When Art is taught, teachers rely on that universality of perception to teach: "don't you see your work is sad instead of merry?" and you see.

For an AI to learn that by itself from data it would need to be able to feel. Yeah, we can attach tags to images and say this one is merry, this one is sad, and then from all of them ask the AI to produce something that matches that label.

Now, another thing of human Art is it seldomly expresses one emotion, but many together in a "harmonious" way (one that matches our inherited, experience-derived, learned and unconscious expectations). Without that background an AI cannot produce something that complex except a) by casuality or b) because someone told it all the details, which emotions/labels were required and in which proportions, in all the complexity. But then, it is not the AI making the Art, the artist is the one who verbalizes all these details (in prose or poetry).

Again, all boils down to "what is Art?" If we do not know how to clearly define it, we cannot make a device (like an AI) that does it.

I put the stress on emotion, for we know most living beings respond to some form of art: flowers, animals respond to music, and we know even fungi seem to communicate in harmonic patterns. Art is not something man-made, is in the way organisms recognize and react to pattern.
 
Your post reminded me of something. About a month ago, I did an image search of Kay Nielsen's artwork because it's just so beautiful and I sometimes need to look at it. When I did, I noticed a couple "off" images in the image search I hadn't seen before. At a very quick glance, they looked just like Kay Nielsen's work, but when I clicked on them and looked at them more closely, I saw they were flatter, more modern looking, and just lacked that certain "something" that can be seen in his other works. I then noticed they were actually AI images done in his style, which led to my discovering a whole collection of AI generated images in his style. While the AI generated images were pretty cool to look at, they also read like knock offs to me. They definitely don't capture whatever it is - that hard to define thing - that makes an artwork sincerely Kay Nielsen's. Part of this could be due to Kay Nielsen having worked in traditional mediums, maybe that makes it harder to replicate via AI, but I'm not sure.
Knockoffs. That's the word. I'll be using it.
 
If I know that's AI right from the start, or I later find out a painting that I looked at was AI, its value to me drops steeply, no matter how perfect it was.

I find I'll often experience that feeling of disappointment as well unfortunately. I think what I would find myself longing to see from an AI artist is a cohesive body of work rather than one-off images. The body of work would reveal things about the AI artist behind it - their preferred aesthetic, subject matter, approach to composition, etc. Rather than using paints, pens, clay, etc, the AI artist's skillset would involve the ability to train a machine to create images that consistently express a personal and/or unique vision. Ideally, if one of the images from this AI artist was seen in the wild, it would be easy to tell it was theirs because it wouldn't be highly derivative of (or literally stolen from!) other artists' work - it would be authentically their vision. I agree, I want to feel that human connection from art, too. I might be more apt to believe that can exist in AI art if I can see the intentionality within the AI artist's work through looking at many of their pieces. For AI art, a one-off image is not enough "proof" to me and so comes off as having less value. At least, that's how I currently feel about it. As this technology evolves, my thoughts and feelings may change.
 
Problem is AI has no intentionality, it will respond to a prompt that contains a lot of detail. And if you change the prompt it will change the output. Much like a human, you may say, but the human made the decisions, not the AI. So at most it may be said to produce, from a vast pool of images it "learnt" an average of the subset selected using the criteria stated by the human prompter. That is not "spontaneous creation", that is more like what a DSLR plus post-processing does, only automated and more complex. From that point of view modern generative AI is like asking if a (glorified) DSLR (or Photoshop) can produce an spontaneous image of art. It is the human who asked for it who did, not the tool.

OTOH almost everybody that has joined a Photography forum has witnessed the loads of new users who believe that having a more powerful DSLR will make better pictures: many of us fail in the fallacy of believing it is the tool that crates the art and provides the quality. Perhaps because it is more convenient to believe a more expensive tool or AI will do the work for us and we will not need to master the Art.
 
Back
Top