AI generated children’s book with ChatGPT, Midjourney involved in art debate


Ammaar Reshi thought it was a fun, creative idea: use artificial intelligence to write and illustrate a children’s book he’s always wanted to do for a friend’s daughter. He only gave himself a weekend for it.

But after completing his project, the 28-year-old is a design manager at a Californian fintech company caught in the crossfire of an escalating public debate: Are artificial intelligence tools a grim reaper for art?

Using ChatGPT and Midjourney, Reshi generated draft text and illustrations that stitched together a story that, as he put it, would introduce children to the magic of AI. Both programs, which are free for at least a trial period, require the user to enter prompts, which they then refine by regenerating images or text.

The end result is impressive for anyone unfamiliar with AI, but often far from perfect: images tend to appear with strange anomalies – in Reshi’s case crooked eyes and 12 fingers – and text created by ChatGPT may have quirks and errors that remind us that this is not the case with AI pretty much Person. Reshi spent hours refining prompts and editing the text generated for the book, and he dismisses criticism that he only had to “press a button.”

He has sold more than 900 copies since releasing his book Alice and Sparkle on Amazon in early December. But one look at the ratings – 60 percent 5 stars and 40 percent 1 star – and his Twitter mentions suggest a growing divide over these tools as the public ponder whether they’re starving the starving artist or whether they’re ethical Everyone.

“The man who made it [this] is neither an ‘author’ nor an ‘illustrator’, but in his bio above he claims that he ‘writes,'” wrote one Amazon reviewer. “Our world becomes a joke.”

READ :  How cleaning data and machine-generated "synthetic" data improves AI performance

AI can now create any image in seconds, bringing wonder and danger

Reshi doesn’t hate the technology, but he understands why some are concerned.

“Any new technology that’s incredibly powerful is a bit threatening to people,” he said, adding, “You see people asking, ‘Is this going to replace my job?’ … This concern – we shouldn’t pretend it’s not serious.”

For example, one of the main complaints about AI art is that some tools seem to have learned from datasets of art made by real people – with real copyright protection – to provide the fodder for their computer-generated creations.

Reshi doesn’t have an answer: “People say, ‘Well, if this model is trained on my artwork and my artwork is copyrighted, is that exactly fair or legal?’ But then I think you’re going to get into this philosophical debate, which is how is that different from human learning? [about] their favorite artist or someone who draws Batman fan art? You could argue that the computer is doing the same thing here.” He adds: “I don’t have a concrete stance on this yet.”

AI has already found its way into the creative world. Last summer, a Colorado man won the State Fair art competition with an image created on Midjourney. In November, the Lensa app debuted a new feature that flooded AI selfies to social media feeds. A comedy robot developed by an Oregon State University professor has begun to learn how to assess the crowd as they play and tell their pre-made jokes. Shudu, the “world’s first digital supermodel”, was created by artificial intelligence and was used in a Louis Vuitton advertisement.

Some high profile creators have made it clear their disdain for this technology. Australian singer Nick Cave recently called ChatGPT an exercise in “replication as travesty” – and a song he wrote in his style “a grotesque mockery of what it is to be human”. During a presentation on artificial intelligence, famed animator and Studio Ghibli co-founder Hayao Miyazaki described the technology as “an insult to life itself.”

READ :  Artificial intelligence will not go away

ChatbotGPT worries teachers trying to expose AI cheating

Artists have also banded together online to stage a digital protest against AI-generated art. For the past month, many have railed against the ArtStation platform after AI-generated images surfaced on its website. One protest image implored AI users to “pick up a crayon like the rest of us did”.

Earlier this week, a US law firm announced a class-action lawsuit against Midjourney, Stability AI and DeviantArt, alleging that “billions of copyrighted images” were used in a dataset “without compensation or consent from the artists.”

“AI image products are not just an infringement of artists’ rights; whether they intend to or not, these products will eliminate the ‘artist’ as a viable career path,” says a press release from the law firm Joseph Saveri. It added, “If streaming music is legal, so can AI products.” The law firm did not respond to The Post’s requests for an interview.

Nik Thompson, human-computer interaction expert at Curtin University in Australia, said he’s heard of cases where a real artist’s signature has appeared in an AI-generated image and that the creators are “rightfully upset”.

“The thing is, the cat is out of the bag and there’s no going back, so I don’t think litigation is going to stop these platforms from continuing to evolve and collect as much data as possible,” he said. “It will continue.”

Thompson believes that many overestimate the current state of development of AI programs like ChatGPT or Midjourney, both of which were released last year. Artificial intelligence is really just “a simulation of intelligence,” he said — it can’t think like a real human.

READ :  Datametrex Receives $1.6M P.O. from a Silicon Valley Tech Company

“Over time, we’ll find that it’s not as fantastic as it may seem,” he said. “…I would like to believe that the discerning consumer who appreciates art and creation will still be able to notice the difference and be interested in the work of creators.”

After an explosive backlash on Twitter, Reshi “collected himself” before sharing his latest personal project with the public – a fictional animated Batman video he put together using an edited version of a script he created on ChatGPT. He generated images on Midjourney, scaled them to larger resolutions using AI functions in Pixelmator, and then recorded himself doing a voiceover, which he edited with an Adobe AI tool. He edited the video in the Motionleap phone app.

“I’ve seen claims that this will replace storyboard artists,” he said. “I actually don’t agree with that attitude.”

Although he admits that he may be overly optimistic, he hopes that professional developers can also find uses for these tools. Storyboarding artists or illustrators could test their ideas by generating them with AI and then use their hard-earned skills to create a more sophisticated product, he said. Amateur creators could also use these AI tools to realize their visions, like he did with his Batman video, he said.

As it stands, some amateur video game developers have started turning to Midjourney to generate game content and graphics, while others have used the program to develop visuals for an indie board game.

“A lot of people see this as empowering a new group of creators — the kids who couldn’t illustrate or write a story that were as good might now get a head start or a leap in that direction,” he said. “I see that as a balance in many ways.”