Artificial Art
Are visual artists about to be replaced by artificial intelligence?
What is art? Does art require a human to create it? Is “art” created by an artificial intelligence still art? What do people buy when they purchase art?
For most of my lifetime, these questions have been somewhat academic. Artificial intelligence couldn’t actually produce art. So these questions were largely theoretical.
But not anymore.
Let me introduce you DALL-E. The latest project from OpenAI. DALL-E can take a written text description and create art from that description. Or, at least, it can create creative images from a description. I’ll let you decide for yourself if they truly constitute being “art” or not.
But I can say this: If a human had posted the images below to Instagram or a website, I would not have questioned that they were real art done by a real artist. And a creative one at that.
DALL-E works by accepting a sentence and then interpreting that sentence into its artistic rendition. So I’ll post the sentence first, with the image underneath.
Input: "Male and female scientists in love with each other instead of researching, by Norman Rockwell."
Same input in another style:
Input: “a wise cat meditating in the Himalayas searching for enlightenment”
Input: “Human and AI fall in love and create the future children of the milky way galaxy”
Input: “A human basking in the sun of AGI utopia”
Input: “A rabbit detective sitting on a park bench and reading a newspaper in a Victorian setting”
Input: “An elephant tea party on a grass lawn”
Input: “Transformers dancing at a rave on the moon.”
Input: “A sea otter in the style of ‘Girl with a Pearl Earring’ by Johannes Vermeer”
Input: “An ibis in the wild, painted in the style of John Audubon”
What does this mean?
We are in new territory.
I have to admit to being shocked at how good DALL-E is. I thought we were still years, perhaps decades, from this kind of ability in artificial intelligence.
The algorithms that power DALL-E (known as GPT-3) are already disrupting written industries too. It’s now common for programmers to get large amounts of help from AI, at least in the form of suggestions. GPT-3 has written shockingly good poetry and news articles. In some cases, it seems to practically pass the famous Turing Test.
What are the ramifications of this technology for fine artists? What are the ramifications of this technology for commercial artists? For illustrators? If I need an image for my latest book, blog article, whatever, do I even need to hire an artist? I can just keep working with DALL-E until I get an image I’m happy with. Why would I go to the time and expense of slowly dealing with back and forth communication with a human when I can try out a few dozen or hundred ideas just by typing a few sentences?
We are setting aside for the moment that DALL-E can’t “paint” on canvas - because that’s an easy problem to solve. The hard problem of text to creative image appears solved.
Back to the original questions, which are probably the most important to our audience: what is fine art? Is DALL-E producing fine art? Will people be more interested in buying art from DALL-E than from humans?
Here’s my opinion: Aside from a few technophiles and, of course, the initial interest into AI generated art, I don’t think so. I think people will still want to purchase art that is created by humans. I do, however, see this technology drastically disrupting commercial illustration and stock photo companies.
But I’m talking about fine art now. Art for art’s sake. While this technology may hurt some industries, it may also be a boon to some creative fine artists. Imagine being able to sit down and “try out” 1,000 ideas just by typing them. I can see art becoming co-created by humans and AI. Just like how some code is today created that way. Even word processors today already “assist” writers with grammar corrections. Those will grow to suggest not just sentences, but entire paragraphs. Entire written pieces from different angles.
In some ways, this new technology is no different any new technology. Cameras certainly didn’t kill fine art - they added a new technology that expanded what humans can do with fine art.
In the end, I still think fine art will continue to be ruled by humans. At least by those humans who understand art is about more than just the images they create.
Here’s the important part, so please pay attention:
The reason that humans will still drive most art is that, for the most part, people don’t just buy the image. People buy the story. People by the connection with the artist. The human connection.
We’ve been saying this for over a decade. I don’t know if most artists believe us. But, frankly, you better all start believing us. Because if art is just about the images, you’re toast. You’ll never keep up with a machine that can happily spit out thousands creative images a day. The story is the important part.
If you don’t believe that, consider that people are buying NFTs in massive numbers. People are paying good money for images they can download for free just to be part of the story. And they are just keeping them in their computer or phone, they aren’t even hanging them on a wall in most cases. Why? Because they are literally buying their way into the story of that human artist. That is a human connection.
And that’s good news, because, if it really is about the story, the connection, the feeling AND the image, then human artists still have the upper hand. At least those artists who understand that their story is important.
This human-first way of marketing is what we teach artists in the Sovereign Artist Club. If you want to learn how to market your art in the 21st century in a story-first, human first way. If you don’t want to be replaced by a machine, then consider joining the club for less than $1.50/week by clicking the button below.
PS - This is a big, diverse, important topic. While normally we limit comments on these articles to paid club members, we’ve opened commenting on today’s article to everyone because we’d love to hear diverse opinions about this important topic.
You can comment by clicking the button below:
I think people may be forgetting that the ultimate Turing Complete Computer is the human brain.
The article is excellent persuasive writing.
The new algorithm will be revolutionary in making Art and who buys it. The traditional artist will likely not disappear, but they'll face increasing challenges in selling their Art.
I agree that humans seek connections with other people, being social animals. But they first seek relationships with the self. Their inner world, if you like. If that's the case, and I don't see any way it can't be true, then the first reason someone buys Art is not to connect to other people but to connect to their inner world or self.
Suppose someone is a member of a religion and sees an image painted by an artist who's also a believer that celebrates their faith. There is a high probability that the person will buy the Art if the painting's esthetics are compelling. But if the image makes a mockery of that person's faith, regardless of its esthetic value, the person will likely not buy it. In the first case, the buyer self identifies with the painting and welcomes the connection with the artist. In the second case, the buyer self fails to resonate with the work and resents any relationship with the artist. The point is that the Art must first be a metaphor for something important to the buyer. The connections to the artist are secondary.
My lifelong friend I've known for over 35 years loves to fish. He doesn't enjoy camping and is reluctant to enter forests. He has paintings and carvings of fish and boats on the water. He values that I'm a painter but doesn't like my work. I paint primarily woodland scenery. He says the paintings are terrific but remind him of all the bugs he hates in the woods. My work doesn't connect to his inner world, so he doesn't like it. But a wood fish carving from an unknown artist does, and he buys it! He doesn't accept my work even though we have a 35-year connection. My friend is an excellent example of how the link to the artist is secondary, and the Art must first be a symbol that allows the person to enter their valued inner world before they buy it.
The article mentions how photography was beneficial, but it was also harmful. It was especially true for a portrait artist, illustrators, and other figurative painters. The world had increasingly little use for their skills since the invention of photography in 1839.
But painting survived by reinventing itself. Maybe this will be true when a bot paints a concept in the coming years. We may not be all toast, though some might be.