There was a recent discussion about “What is Art”. So, what do you think of all the AI-generated “art” that is now all over the internet? There are probably many answers to this; for example either it is NOT art, and will never be a genuine, personal, cultural/historical/geographically- located expression of human creativity. Or that maybe AI “art” is the start of something really new and interesting. Remember the critics of Bob Dylans electric guitar, who didnt see the possibilities of electronic music? What do you think? Artists opinions especially sought, please.
1 Like
I’m not an artist, but I am a techie 
and my vote is “not”. As regards electric guitars, it’s still a person doing the plucking. So, I’ve no issue with electric brushes or 3D printers or whatever tools, but the inspiration, for me, should be human, not some random (or not so random) thought generator 
1 Like
Depends on the criteria one uses to define ‘art’. If one chooses to regard AI as just another medium, like photography, then it comes down to intentionality of the person who initiated the work. As I’ve posted previously, putting something in the category of art because that’s where it fits best is a judgement of ‘category’ not ‘quality’. ie it can be ‘art’, but that doesn’t mean it’s necessarily art of quality (incidentally, I think that last phrase is a much more difficult one to consider).
Most of the egs I’ve seen (in reproduction) seem closer to illustration than to art, because they appear to do little more than illustrate, parody or plagiarise. However, one could argue that AI generated art has much in common with certain forms of Conceptual Art from the 1960s which just consisted of a typed series of instructions that the spectator could follow if they so desired. It seemed exciting at the time (at least to some), but it hasn’t aged well.
One of the several limitations of current AI generated works is that, unlike paintings they have a ubiquitous finish because they are all printed in the same way. A second difference is that most paintings are built up through layers of underpainting and correction of previously unexpected visual problems that need to be resolved in order for the painting to progress. This evolutionary exchange of dialogue between artist and medium is not apparent in AI generated work (though some internal process of acceptance /rejection of possiblilities has occured at an invisible, computational level).
I think AI generated work might be more interesting and less derivative if output through 3D printing, which also allows a much greater choice of printing medium - from plastic to unfired clay for instance.
Could write much more but have to remove a chicken from the oven. Thanks for starting this thread!
2 Likes
The ‘inspiration’ (don’t like the term but I’ll stick with it) remains in the human realm because the process of producing the image is initiated by a human. There may also be a post-production editing of different outputs. Rather than think in terms of comparisons between art and music, it might be more use to think of a comparison between sounds/noises and 2D visual imagery which are much broader categories.
As a guitar player (gigging later this evening) there’s plenty of acoustic guitars being used in very formulaic, uncreative ways. What counts as creativity and what is just going through the motions? It can be hard to be creative every time you go out and play the same set, more or less.
Is ai generated output art? Probably not, but the use of ai and it’s output might be.
IMO of course.
1 Like
Yes, I’m sure the term is incorrect, but it was all that came to mind.
Is it though? I have to think about that. I can understand an AI system scanning X-rays and using rules to identify tumours better than the naked eye. But an AI system scanning a sunset? The image is run through a whole bunch of data driven rules and then… what? The human input has been in writing the rules, if something new and weird and wonderful pops out, is it the rule writer that has created it?
It might not be ‘programmed’ as you think of it, but the computer programme is given a set of parameters by a human, and of course as you rightly note, there has also been ‘human input in writing the rules’.
I think in the majority of instances the output will be banal - because the human input will also be banal, OTOH if the input is creative in being able to use the AI generative process in an unexpectedly creative manner, ie. going beyond illustration, but using the possibilities of the process to say something, then that’s a different matter. It’s a question of going beyond mere appearances.
I think you’re referring back to another SF thread on photography and art -the same criteria and distinctions apply - not all, in fact most images aren’t, they fall in other categoris of 2D visual phenomena, but like photography AI is another medium for producing really strong art, but like any other art making medium, the quality of what is produced will depend partly on the quality of the artist/producteur and partly on thier understanding of this new medium’s possibilities.
PS hope the gig goes well.
1 Like
David Hockney + iPad? Definitely art!
Elephants painting art?
As long as it doesn’t involve cruelty to the animal, why not? At least an elephant is sentient, unlike AI.
So far
Once again, for me at least, there’s a conflation of initiator(artist) and medium (elephants , paint and paper).
For about twenty years or so, I’ve been aware of this pair who IMHO ‘exploit’ animals as an art and money generating medium - there’s a false, but commercially appealing notion of some sort of ‘authenticity’ if the mark/s on the paper are made by a wild animal
Incidentally, apropos painting, adult elephants, like rhinos are virtually blind, but have a phenomenal sense of smell.
On a dawn bushwalk in a Swazi game reserve (with an armed ranger) I came within about 30m of a familly of rhino in open space, but only the calf could see us and didn’t emit any danger signals. Memoraable experience to be in the animals’ space
I believe with Hockney, he was actually inspired into new avenues of depiction by the newly creative iPad medium.
I am not sure about the elephants and only hope they enjoy waving the brush and paint about. Endearing results but not so sure it qualifies in the grand discussion as art.
David Hockney was originally an etcher and he continued working as a printmaker during his postgrad at the RCA (my alma mater too!) . So, though best known as a painter, since the mid 1950s he has continually explored more technical means of image making - etching, lithography, dyed paper pulp casting, usually working with master printmakers who are experts in their medium.
About thirty years ago there was a series on UK TV that included Hockney making his first computer generated images via a technician who knew how to use a Quantel Paintbox which at the time was the latest thing in computer generated image making, because you didn’t need programming skills to generate an image. Around the same time (1989) I did an in-service course in ‘Advanced Computer Paint Systems’ at Mx Poly, which had one of the UK’s four Quantels. As I remember a Quantel cost around £45K but within a couple of years its keyboard and graphics tablet based communication had largely been superceded bythe mouse on Mac based Photoshop and Windows based vector graphics animation programmes like 3D Studio. Also suddenly there was sufficient memory for a much welcomed ‘Undo’ facility.
So within a couple of years the situation changed from a traditional one similar to that of an artist to a Master printmaker - Wikipedia to one where the artist could handle the entire process. Financially successful artitsts, began to use the new digital editining studios to output their work as giclées (hi-res archival quality inkjet prints). However for me giclées unless in the now seemingly defunct Quadblack process have a ‘dead’ surface compared to traditional fine art etchings, mezzotints, or even lithographs. And, to come full circle for me this is one of the current limitations of the new AI generated images -a boring, neutralising surface
There is also the current issue now of artists suing for copyright when AI uses their works without permission or acknowledgment. According to recent news, there are many legal challenges coming up from artists and image organisations;
" Three artists filed a class-action lawsuit in San Francisco against AI art tools built by Stability AI, Midjourney, and DeviantArt. Stability AI faces a second case filed by Getty in the UK.
“AI image products are not just an infringement of artists’ rights; whether they aim to or not, these products will eliminate ‘artist’ as a viable career path.”
“A statement from Getty Images isn’t quite as sour on AI, suggesting that “artificial intelligence has the potential to stimulate creative endeavors.” But Getty’s legal claim is similar. Stability AI “chose to ignore viable licensing options and long-standing legal protections in pursuit of their standalone commercial interests,” the company said.”
Have you tried a print on aluminium. I have had a couple done by SAAL in Germany, and the surface had a distinct texture from the printing process that’s quite unlike any conventional photo print and gives the surface a life of its own. The mono print is especially good, with a real sense of depth, though I’ve had more comments about the colour image where it’s sat in my office.
Thanks, but it is very different, the prints I posted on the link below were printed on aluminium, and it’s great for large prints, but the surface doesn’t give anything back, unlike Quadblack pigmentYour photos - the view from your house - Hobbies/Arts/Crafts / Photography - Survive France
Here is an example of the human input in an AI program; basically you use language to “extend” an existing work or style. For example, you ask for a painting of the Taj Mahal in the style of Vincent Van Gogh’s “Starry Night”. Of course, this is a DERIVATIVE use of the original, not an authentic human re-imagining (like for example Dali did). It could nevertheless be interesting, and in the current favorite language, a “re-imagining” of classic works of art. I predict it will be used by museums and galleries to refresh interest in the classics, in the same way that large scale light projections of art like Van Gogh and Monet have been used by galleries and art spaces to attract new and younger audiences and keep themselves “relevant” to a modern audience. However, this kind of art relies on the cache of the ORIGINAL works; the value of art is predicated on the rare, original work, and these re-interpretations can be seen as similar perhaps to prints and posters, calendars and teatowels, screensavers and T-shirts. The most popular artworks and museum objects in the world are basically “icons”; the Mona Lisa, the Rosetta Stone. Humans have an intrinsic need to see/touch/commune with the “real thing”; this need is almost a religious hunger. This hunger for authenticity in this modern world of limitless reproduction and facsimile may, in the end, even INCREASE the value of human-created art. Food for thought.
I would agree that the real thing needs to exist first and foremost. Although, art no longer seems to demand original be one of a kind.
At the same time, in commercial production there is no accounting for taste, so endless possibilities for more tchotchkes. These appeal precisely because they evoke the original, famed and lauded as Art.
As fascinating as the developments in AI creations may be, for me, it is without craftsmanship. In wanting to be close to an original, I feel it means being able to see the hand of man.
1 Like
Agreed. Interesting but not art in itself.
Let’s see what artists do with it. Possibly, in irony.
Love all you say on Hockney. A curious and innovative fellow. I expect he really enjoyed the backlit effect if light making his artwork colours come alive.
This is right up my street! One of my main hobbies is photomanipulation and from experience I can say that it takes quite some time to master the main programs available. Virtually all my photographic output is cropped, sharpened, brightened, recoloured to achieve an image more to my liking. There are endless possibilities to alter an image from replacing the sky to removing unwanted people or cars. Then there are countless ways to distort images with filters to produce creative images. The important thing to note is that the vast majority of these alterations are possible without any intervention by artifical intelligence but skill is needed to use the popular programs and you absolutely need these skills in order to make the best use of AI.
This is the process for creating the attached image:
Background – merging of two separate Dalle-2.AI images from a text prompt, with extensive inpainting and outpainting. Perspective was added by the repetition of the street lamps and by adding the edge of the path on the left. Almost all of the colouring on the left and the merging of the two original AI images was done by hand
Figure 1 – Wombo AI (completely recoloured, resized and partly redrawn to match the other person.) Figure 2 - AI Dalle 2. Needed some redrawing/recolouring/resizing.
The image was tidied up in AI Deep Dream Generator and Filter Forge.
The AI images are never perfect. Some programs cannot reproduce human or animal faces. A simple request resulted in a woman with three legs. AI images even at best are unpredictable and quite often
do not produce a usable image. The artist’s vision and ability to mould the images into an acceptable new image are paramount.
On the other hand, “artworks” produced by artificial intelligence and zero human intervention are a completely different case and are not to my taste.
2 Likes
I agree, Susannah, that the word “craftsmanship” is at the beating heart of the meaning of art. Before the 17th century the word “art” referred more to skill, craft and the mastery of a medium, more recently the word has been enlarged to encompass a much larger field of meanings including creativity itself, aesthetics, an attitude, a way of seeing, “taste”, “high art” as opposed to mere crafts. I think the history of this change has to be seen against the background of the industrial revolution and the mechanisation of art. So AI can be seen as just the latest version of the mechanisation of art, which once again, strips the individual from his/her creation (Marx was spot on) in order to get the most reproducible value out of individual, creative labour.
Given all this, my first reaction to AI “art” I have to admit, was disgust, which was why I asked the question of what others thought of it. However, what disgusts, also fascinates. I’m sure some people will be inspired to wrestle with, or come up with an oppositionary movement to AI “art”. Whether working within and with it, or in a movement against it. A new “Arts and Crafts” movement? Perhaps, as Mark said, a new kind of Conceptual Art? Or will the younger generation, so adept at technology, wrest the codes and the reins from the machines and create a whole new oeuvre of human-centered AI “art”?