I studied photography and sculpture, and initially the only thing that made any difference to whether I enjoyed a piece of art was interpretation by someone who's seen a lot of art. There were only two or three people who could do this effectively for me. Eventually, after talking to them a lot and seeing a lot of art, and thinking about it deeply, you start to develop your own tastes, and realise what kinds of things speak to you.
This project seems under the misapprehension that there's some kind of "perennial philosophy" of art that can be uncovered by asking enough people about randomly generated images. For one thing, this is absolutely not what I (and, I would argue, pretty much anyone who's spent time thinking about it) consider to be "art". Also, the experience of art is not just pretty pictures, but all of the past experiences that you bring with you when looking at a piece of art. Even people who like what I would consider to be twee pictures do so because of what they're bringing to the aesthetic experience.
If statistically significant results come out of this, I'd argue that any information it generates is going to be extremely shallow. It's like a kind of statistical fetish.
Of course art is much more than randomly generated aesthetics, but I still think this project can be insightful, if not useful. Sure, it's not telling us what objectively makes great art, but I don't believe anything can.
But if statistically significant, it might tell us something more fundamental about the pure aesthetics assigned by people to patterns. It's more like the bouba/kiki experiment.
As I said elsewhere in this thread, I think this probably just approximates a description of the neurobiology of vision, from the retina to the occipital cortex. For example, newborns prefer high contrast images, but I presume no-one would consider them to have any understanding of art. The same for things like red / green preferences, which are likely to have their roots in the red / green cones in the retina. I'd argue it's not just useless, but lacks any meaningful insight into art.
I agree it's not particularly useful as far as art goes, but I think it can be insightful as far as visual aesthetics go.
The mistake to me is in implicating art, instead of just aesthetics. But Art Vote is a lot more catchy than Neurobiology Of Visual Aesthetics From Retina To Occipital Cortex Vote.
It's sorely obvious that the creator of this project was a computer scientist, or someone who thinks in an overly technical manner rather than an artist, who thinks more about emotions and feel.
I had a very different reaction. I had the sense this was an art world person (student or otherwise) who wasn’t technical enough to make it interesting. Many computer generated art projects are much more dynamic and produce much more passable art (https://ml.berkeley.edu/blog/posts/clip-art/). That said, I think I would have to agree with other comments here, I don’t think any deep conclusions could be drawn from these results. For starters the artworks are really repetitive and i would guess it biases respondents quickly toward the negative.
> I studied photography and sculpture, and initially the only thing that made any difference to whether I enjoyed a piece of art was interpretation by someone who's seen a lot of art. There were only two or three people who could do this effectively for me. Eventually, after talking to them a lot and seeing a lot of art, and thinking about it deeply, you start to develop your own tastes, and realise what kinds of things speak to you.
I think that might be specific to you, or at least not how everyone appreciate art. My enjoyment of a piece of art is directly linked to how I feel when I see it/think about it. I can enjoy something without explanations and not enjoy something with explanations. I don't think I'm in the minority here.
> If statistically significant results come out of this, I'd argue that any information it generates is going to be extremely shallow. It's like a kind of statistical fetish.
They will be, but not because of the nature of the project, but because of the implementation. There are only a few variations of a few basic things.
> This project seems under the misapprehension that there's some kind of "perennial philosophy" of art that can be uncovered by asking enough people about randomly generated images.
I'm not sure this is a misapprehension. The project itself is limited, but there are information to learn with randomly generated images. If people like art A and not art B, you could generate a graduation of A to B and see if people like art close to A (which could mean they like art "like" A), or if they only like A (which could mean that A has something specific).
Live stats at web scale is a really fun problem to have. Un-optimized, using public cloud infra, round-trip latency for collecting data, calculating statistics, and render/viz to the user, it's shockingly slow! For crowd training Art GANs, latency probably doesn't matter, but seeing the live stats displayed is still magical to me. Imagine how fun ArtVote would be just using the public collection of images from MoMA, seeing the crowd evaluate a new work live ;)
I sort of love projects like this (i.e. the attempts to automate humanity itself by trying to naively quantify experiential subjectivity) exactly because they force the issue. Each silly example of "oh look, a machine can paint, make music, write stories" gives us an opportunity to say: that's not the point.
I don't see what this could possibly tell us about art. The generated images are both too random, but also have absolutely no character or variation.
There are so many brilliant sources of understanding on what makes art appealing, and so many images of art on the internet. I don't get why the site would use these drawn-in-ten-seconds-in-mspaint images for voting.
I think that in principle this test could produce interesting and useful results. Having researched in the field of computational aesthetics, I speak with some experience. Much of the aesthetic value an artwork 'lives' in such abstracts as its 2D geometry, its 3D geometry, its color contrast etc. I would have preferred that this test be designed to test just one of these attributes.
However, even on the currently small response rate, there seems to be some valid patterns emerging...
In the Overlapping Drops results, the red/green images is the clear favorite. This is no surprise, as this antagonist pair is the most favored in art. Around 90% of the paintings in any national collection are painted along the red/green axis.
Neither am I surprised that the most popular Nested Squares results is the one that features the greatest contrast. 'Higher than nature' contrast is a core requirement of art. This is as true in Korean soap opera as it is in visual art.
I believe that the experiment could have been better designed. A better intro to the experiment would have been a good idea. Also, (in my opinion) I believe that the images should have been presented as pairwise comparisons (Thurstone, L.L.: A Law of Comparative Judgment) with a few repeats shown in. The preference choice should be restricted to three from five. This would have made it easier to get more responses from each user. Perhaps it would have been better to have kept the images in their family groups. In that way, it would be more clear what is being tested.
I don't want to be too harsh, but in my view this is mostly tosh. For example:
> 'Higher than nature' contrast is a core requirement of art
There are so many counter-examples I don't know where to start. Turner's watercolours. Hiroshi Sugimoto's seascapes. Do So Huh's installations. I mean come on, what about Rauschenberg's White Painting? I suppose if you don't consider anything after the 18th century you might be right, but it feels like you're deliberately narrowing down what you call "art" so you can make confident statements like that.
Sure... to those counter examples, I could add more: Gwen John's muted pallet, Joseph Albers' same-hue combinations,
It is a trivial task to find exceptions to any principle in art. However, as a general rule, it is true that artists exaggerate. My own work examined 100 American paintings all painted at approximately the same time, of similar subject matter. These were compared to the same number of snap-shot style photographs of similar subject matter.
The results showed us that artists exaggerate the strong color differences of nature, yet bring weak differences closer together. Hence an artwork is (in general) an exaggeration in two dimensions: contrast and affinity. This is done differently in the hue, saturation and lightness channels, with (as you might expect) lightness featuring the strongest. However, there were interesting exception in the hue treatment of particular subject matter.
> I suppose if you don't consider anything after the 18th century you might be right.
You have a point, if you only consider fine art, and your examples show that you know your businesses in this domain. But I would say it is even more true of such applied arts as game level design, high vfx fantasy movies, Deviant art style fan art and such like.
I also have a degree in experimental psychology and a masters in neuroscience, and to be honest, I think you're just describing the neurobiology of vision. Newborns prefer high-contrast images, but I presume you wouldn't consider them to have an understanding of art. And the "red/green" axis just makes me think of the red/green cones in the retina. It sounds like a very naive view of "art".
The comment "if you only consider fine art" also feels like you're moving the goalposts, where you're saying "within these works of visual art people prefer high-contrast images", but the categorisation of those images is that they're a subset of visual art within which people prefer high-contrast images.
I am also sure that I am describing the neurobiology of vision, though I would love to know more about it. Such preferences need to come from somewhere. For example, I have always thought that the popularity of the vignette derives from the fact that our visual field is oval not square.
As for the fine art issue…. As the previous poster implied, art pretty much changed form around 150 years ago. It went from being pretty much rule-based to pretty much rule-breaking. It would be a waste of time to apply My own research on art to (for example) Tracy Emins tent or Joseph Beuy’s dead hare.
I always thought that art would be much easier to talk about if we came up with different names for the stuff that was made since Picasso’s gurnicca and the the stuff made before. They are different things, with different functions and it is only confusing that they are both called art.
I used to run a DA like website in 2005 and at one point had to categorize all uploadable art. The truth is art is in the eye of the beholder. There is no way to actually say something isn't art if someone sees anything as such.
100% agree. I think "the beholder" is what is often overlooked. Marcel Duchamp submitted a urinal as a piece of art (an act of provocation).
> When explaining the purpose of his Readymade sculpture, Duchamp stated they are "everyday objects raised to the dignity of a work of art by the artist's act of choice."
The important point is the involvement of a person. Computer generated art can't be art unless someone thinks it is.
I actually think that if the intention of the original programmer is to say "anything created from this code is art no matter what it is" then it's basically impossible to say otherwise.
I feel like this discussion devolved into "what is art?", while completely missing my original thesis. Of course every image this app produces is art. But that's the point... art is infinitely broad, and the art that this app produces is trivially narrow. So this isn't an argument about art, but instead well-designed surveys.
The only possible conclusion one can draw from this app isn't "what are people's art preferences?", but instead "what are people's drawn-in-ten-seconds-in-mspaint preferences?"
The app would just as accurately gague people's art preferences if it presented only photos of shoes.
I feel like this discussion devolved into "what is art?", while completely missing my original thesis. Of course every image this app produces is art. But that's the point... art is infinitely broad, and the art that this app produces is trivially narrow. So this isn't an argument about art, but instead well-designed surveys.
The only possible conclusion one can draw from this app isn't "what are people's art preferences?", but instead "what are people's drawn-in-ten-seconds-in-mspaint preferences?"
The app would just as accurately gague people's art preferences if it presented only photos of shoes.
I don't think the algorithms produce good art. Aside from that, I think changing the rating system from an emotion-based one to something more neutral like numbers would be better. Art that makes you feel "bad" can be good art.
The rating scale should be quantified in that case. But yeah, I always felt that way too. Pretty landscapes and stuff end up going mainstream but I've always felt "true" art in any medium has to have something slightly jarring or uncomfortable about it. And I'm not talking about jarring and uncomfortable discordant random lines and tones made by AI...
I like the idea but the way the pictures are generated seems way too obvious. It's clear that there is a small set of "patterns" with randomized parameters. So it becomes like "do people like the water drops in the corner or in the center? red or black?". Might be more interesting if the generated images were a little more sophisticated and varied.
"In 1994 the artists Komar & Melamid commissioned a public-research polling firm to conduct The People’s Choice, the first poll on artistic taste in the United States. Individuals were asked approximately 100 questions on a variety of subjects, ranging from their consumer tastes and recreational activities to their knowledge of famous artists, and their preference for angles, curves, brushstrokes, and particular colors, sizes, content, and style in painting. As Russian émigrés, the artists were intrigued by the idea of the consumer-research poll as an outgrowth of American democracy. At the same time, their interest in democracy led the artists to ask what a genuine people’s art would look like. What is a democratic and populist painting?"
Low-contrast images will always be unanimously down-voted (for good reason: us humans can't even see the art clearly). Wonder why the algorithm doesn't use a sane heuristic for picking color palettes.
There might be a way to create algorithmic art that has meaning. But looking at these images, I am not feeling much. They look too much like other graphics I have seen already.
If anyone is interested in understanding more about art and human perception of art, then this book is as good a starting place as you'll find: https://en.wikipedia.org/wiki/The_Story_of_Art
As the book says, ″There really is no such thing as Art. There are only artists.″
Every image it generates is one of 8 “pieces”, where each “piece” is an algorithm that has some randomness but a predetermined basic design. [1]
My impression is that the instantiations of each “piece” tend to all look the same and have similar aesthetic value. Personally I like Nested Squares, 45 Degree Paths, and Bezier Curves when it chooses rainbow colors. On the other hand, Overlapping Drops is fundamentally ugly (looks like a 5-year-old using a paint program’s stamp tool) and Patterned Lines is usually ugly (too many lines, too much contrast, no antialiasing, reminds me of the Windows pipes screensaver). But I feel like both the good and bad come mostly from the human who wrote the program, not the random generation itself.
There are clear tendencies there regarding colours, contrast and complexity, and they do provide some (limited) insight into what viewers expect in art.
Disagree with you, there are no clear patterns other than "slightly more interesting than the alternative". Most of it can likely be chalked up to being noise in the sampling process.
This is too random to be meaningful for "art". But it might be able to build some stats on design layout choices... and if it does so, comparing its results to the general design philosophies built over the years could be an interesting exercise.
This project seems under the misapprehension that there's some kind of "perennial philosophy" of art that can be uncovered by asking enough people about randomly generated images. For one thing, this is absolutely not what I (and, I would argue, pretty much anyone who's spent time thinking about it) consider to be "art". Also, the experience of art is not just pretty pictures, but all of the past experiences that you bring with you when looking at a piece of art. Even people who like what I would consider to be twee pictures do so because of what they're bringing to the aesthetic experience.
If statistically significant results come out of this, I'd argue that any information it generates is going to be extremely shallow. It's like a kind of statistical fetish.