I would prefer the term in the title, “Computational Art”, but it’s going to be hyped as “AI” Art no matter how silly that is. At least in its current state it’s not “intelligent “ at all.
But it’s interesting. Ars Technica just did an article about an iOS app that will generate “art” from a description typed into a text box.
As the piece warns, it can generate NSFW imagery and therefore might get removed from the App Store at any time. Get it soon if you’re curious.
My results so far have been bizarre for the most part. But maybe there’s a trick to entering a good description. Here’s one I got, the best of three attempts using the same description each time and clicking on the seed number to get differing images.
Is it Art? You decide! Curiously, my iPhone 13 Pro took five or six image generations before it felt warm and I was using 512 x 512 output. The writer’s iPhone 11 Pro got hot right away and gave a warning, then crashed, on the 512 x 512 setting. Maybe the app got an update since then?
Sorry to continue my dystopian ranting around here this morning, but whether computational art or AI Art, it is FUBAR. We are losing what it means to be creative human beings…the winning artist’s 800 hours of phrasing his artwork description in the computer software gives rise to a real life implementation of the infinite monkeys theory…
It also reminds me (too much) of Arthur C. Clarke’s short story Nine Billion Names of God
There’s a blog that often shares this sort of stuff. Gurney’s Journey.
As an artist myself, the real issue is that what the AI disrupts is the idea of an artist as a commercial venture, where you illustrate a magazine or advertisement or such. Those examples are all about the lowest common denominator. The client just wants an image, and they don’t necessarily care who (or what) produces it. We live in a culture where that’s just the truth.
Alternately, what AI can’t take away is the experience of making art. And the desire people have to purchase art from someone that talks about their shared stories and spaces.
I wonder if portrait artists felt the same when the camera was created?
The current generation might still value buying art made by human, but I’m worried the next generation who grow up with AI might find drawing art “just a hobby for people who have too much time, and not a real job”.
It’s truly disheartening for those who spent decades to perfect their art.
Many of those people keep claiming: “it’s a tool for artist to draw faster, they just need to edit and perfect the drawing, just like photobashing. Just embrace it and you will win.”
It’s no where the same. Many people pride in making their art from scratch. I’m not comfortable calling something 90% not made by me my own. And while AI still requires human intervention to look perfect, the barrier entry is much lower, it’s more of a manual job than actual art. Added the fact that the machine “ate” many copyrighted artworks by artists spendings years making, without consent to get to the state it is now, it’s a pill that’s tough to swallow.
And the AI while is still imperfect and occasionally generate disfigured human, which requires human editing, who know what will happen in 2, 3 or 10 years? A few years ago I visited the technology and it’s no where as good as it is now.
I’m not trying to be contrary, just seeking to understand and build a philosophical framework. Forgive the lawyer instinct to define and categorize.
If I have an eidetic memory, and spend weeks in museums around the world, and/or scouring images on the web, before I begin drawing or painting, did I do the same thing as the AI that “ate” copyrighted works?
Bards and troubadours in the Middle Ages were notorious for tweaking existing material. I suppose a modern equivalent is the number of artists singing a cover of Jingle Bells. My (dim) recollection of Art Appreciation includes dozens of Madonna and Child themed works. I am reminded of the Star Trek episodes where Data painted and played music based on reviewing records.
Is copyright law the right framework for defining this process? Not sure, just thinking out loud. Thanks for sharing your thoughts.
This is fundamentally different in that the human “imitator” or more politely “inspiree” in your examples still controls the ultimate artistic outcome, while this AI magic has mined thousands of pieces of art/photography to composite the outcome.
I think you’re right on point. Striking the balance between creators, distributors and material either already in the public domain, or sufficiently common/generic that it should be is the dilemma.
Mass market corporate distributors really drove the last round of copyright expansion in the US Congress pushing the balance of power towards commercialization. Since those players very frequently buy out individual artists by license or work for hire, those laws do less to help the individual artists.
Walt Disney was a creator. The company that bears his name today consumes creators. JK Rowling is a creator; she’s done a masterful job of controlling the evolution of her original concepts. But the shape of the pyramid applies to creators like it applies to the millions of kids who will be playing basketball this weekend: there are a miniscule number of pro contracts at any given time.