Art by AI: “Scary and Fascinating”

The art of creating artificial intelligence


The art of creating artificial intelligence

06:53

DALL-E 2 is an AI program that can turn anything you write into art in any style. Do you want a picture of a panda in the style of Renoir? Here ya go!

ai-panda-art. jpg
Magnificent!

CBS News


Sugarplum fairy eating a Kehinde Wiley-style cheeseburger? try this! One woman said, “This is both scary and awesome!”

ai-sugarplum-fairy.jpg
A vision of the AI ​​sugar fairy dancing in your head.

CBS News


People use DALL-E to make music videos…


Mr. Blue Sky – But every lyric is an AI-generated image by
solar prophet on YouTube

…plus children’s books and magazine covers.

alice-and-sparkle-cosmopolitan-ai-cover.jpg

CBS News


I even used it to illustrate “Sunday Morning” stories.

ai-cannon-art. jpg

CBS News


DALL-E 2 and its competitors, such as Midjourney and Stable Diffusion, are available to anyone; They are inexpensive or free to use. It’s easy to see how this technology will change the game in graphic design, interior design, architecture, fashion, and filmmaking.

The creator of DALL-E, Aditya Ramesh, works at OpenAI, a company founded in 2015 by Elon Musk and others. “The company’s goal is to develop artificial general intelligence,” he said. “By this, we mean an artificial intelligence that can do all the things a human can, deploying them in a safe way and maximizing the positive benefits to society.”

To train DALL-E (the name is a combination of Dali, the artist, and Wall-E, the Pixar robot), the company fed him 600 million images mapped from the Internet.

“It’s not just cutting and pasting together; her understanding of images is kind of more abstract, kind of like how a human being uses inspiration from all the images that he might have seen in his life,” Ramesh said.

I know what you’re thinking: This will put a lot of artists out of business.

Meet concept artist Carla Ortiz, who has designed characters, creatures, and costumes for several Marvel films, such as “Doctor Strange.” Why would someone hire someone when they could have something [AI] Is this “good enough”? ”

karla-ortiz-at-work.jpg
Carla Ortiz, Creation of a man-made work of art.

CBS News


But what worries her is not unemployment. The work of professional artists is over in the OpenAI database. That’s how DALL-E knows how to imitate the style of Norman Rockwell, Picasso, Ansel Adams, or other living working artists.

ai-norman-rockwell-ish.jpg
Or Norman Rockwell.

CBS News


“It’s an invasion of privacy,” Ortiz said. “It’s an invasion of our ability to consent to be in these datasets, you know. Nobody Requested we. And the way to fix that is to do that by building datasets that are completely populated with public domain work, and then any other kind of expansion of that is done through licensing agreements.”

The idea is that if an artist chooses to “subscribe” to the AI ​​database of images, and it becomes part of its algorithms, the artist will be compensated.

Technical AI companies say they are listening. For example, Stability AI recently announced that it will allow artists to “opt out” of future versions of its database.

But OpenAI is also concerned about other downsides, such as people producing AI-generated images that contain pornography, violence, or misinformation. “When we trained the model, we filtered out images of weapons, blood, and gore,” said Ramesh.

So if someone enters “the president kills cats” into DALL-E’s image generator, it will return an error message. “She will not allow you to do that,” Ramesh said.

DALL-E also attempts to compensate for the stereotypes of race and gender in the online photo world. So even though 90% of pictures of doctors on the web may be white men, Ramesh said, DALL-E will try to straighten things out.

But not all AI companies have built in these kinds of safeguards. According to Imad Mostaki, CEO of Stability AI, “There are a lot of different points of view and a lot of different points of view on this topic. As a society, we have to come together and figure out how best to use this amazing technology.”

ai-stabil-diffusion-cityscape.jpg
An artificial intelligence cyber city from Stability Diffusion.

CBS News


Stable Diffusion software, from Stability AI, is open source, which means it’s free to anyone, without restrictions or firewalls. This approach has set many alarm bells off.

“We think that putting this out in the open so people can see the power of technology and then see how we can mitigate the damage beats being in the boycott of unelected companies,” Mostaque said.

Some Stable Diffusion fans produce harmful and shocking images, but they rarely see the light of day, according to Mostaque, because Twitter and Facebook block them. “If you put it on social media, or post it there, it will be treated like any other bad content,” he said.

The latest advances in the art of artificial intelligence have arrived there. These early apps still had trouble with text, faces, and generating the usual number of fingers.

how-many-fingers. jpg
how many fingers

CBS News


But they are getting better quickly. Meanwhile, AI applications that generate audio and video are already being tested.

For artist Carla Ortiz, these are sad developments. She believes there is value in the creative process itself: “It’s therapeutic. It’s inspiring. It’s human-to-human. AI tools can’t do that yet.”

But Emad Mustak, CEO of Stability AI, specializes in the arts of artificial intelligence. “I think it’s one of the biggest leaps forward in technology since perhaps the Internet,” he said. “It will create entirely new industries, and it will make media more interesting and more entertaining. I think that creates a lot of new jobs.

“It’s coming. I think it’s going to change almost everything.”


for further information:


Story produced by Sarah Coogle. Editor: George Bozderik.

Leave a Comment