One of Pixar's most distinguished animators, Lee Unkrich, was a seventh graders. He is looking at a train locomotive on a computer screen. He thinks it's amazing. Lee learned that the image had not appeared simply by asking for a picture of a train. It had to be rendered by hard- working humans.
Imagine if Lee stumbled onto DALL-E, an artificial intelligence that can generate original works of art based on human-supplied prompts that can be as simple as a picture of a train. The wow is back when he uses words to create images. It doesn't go away. He describes it as a miracle. The results took away my breath and made me cry. It's that wonderful.
We have crossed a threshold. We have been reassured that computers weren't really creative. Millions of people are using new breed of artificial intelligence to generate stunning, never before seen pictures. Lee Unkrich is a professional artist but most of these users are not. Not everyone can write, direct, and edit an Oscar winner like Toy Story 3 or Coco, but everyone can use an artificial intelligence image generator. The realism of the screen is amazing. The universal response was "WOW". More than 20 million images are created every day by humans working with artificial intelligence. Artificial intelligence has become an engine of wow.
The output of these artificial intelligences is dependent on what we expect the pictures to look like. They restructure the new pictures in a way no human is likely to think of, filling in details most of us wouldn't have the skills to do, because they are an alien artificial intelligence. They can be told to make more variations of something we like. Their most powerful advantage is that they are able to make new things that are understandable but also completely unforeseen.
There is a thought that comes to everyone who has seen these new images, that human-made art must now be over. Who can beat the speed, cheapness, scale, and creativity of these machines? Is art something we must give up to machines? What else can computers do that we were told they couldn't?
I have spent the past six months using artificial intelligence to create thousands of striking images, often losing a night's sleep in the pursuit of just one more beauty hidden in the code. After interviewing the creators, power users, and other early adopters of these generators, I can say thatrative artificial intelligence will change how we design just about everything. Humans will not lose their jobs because of this new technology.
With the help of artificial intelligence, images can be created. The best applications of this power are the result of long conversations between humans and machines, not a single prompt. The advancement of machine learning has led to many, many iteration, back-and-forths, detours, and hours of teamwork.
The marriage of two technologies led to the creation of an artificial intelligence image generator. One was a line of deep learning neural nets that could generate realistic images, and the other was a natural language model that could be used to interface with the image engine. A language-driven image generator was created. Billions of examples were used by researchers to connect visual forms to words and forms. Human users can enter a string of words that describe the image they want, and the prompt will generate an image based on those words.
All Rights Reserved © 2024
