AI puts its artistry to work creating new human proteins
“One of the most powerful things about this technology is that, like DALL-E, it does what you tell it to do,” said Nate Bennett, one of the researchers working in the lab at the University of Washington. . “From a single prompt, it can generate an infinite number of designs.”
The Rise of OpenAI
The San Francisco company is one of the most ambitious artificial intelligence laboratories in the world. Here is an overview of some recent developments.
To generate images, DALL-E relies on what artificial intelligence researchers call a neural network, a mathematical system loosely modeled on the brain’s neural network. It’s the same technology that recognizes commands you bark into your smartphone, enables self-driving cars to identify (and avoid) pedestrians, and translates languages on services like Skype.
A neural network learns skills by analyzing large amounts of digital data. By identifying patterns in thousands of corgi photos, for example, he can learn to recognize a corgi. With DALL-E, the researchers built a neural network that searched for patterns by analyzing millions of digital images and the text captions describing what each of those images represented. In this way, he learned to recognize the links between images and words.
When you describe an image for DALL-E, a neural network generates a set of key features that this image can include. One feature might be the curve of a teddy bear’s ear. Another could be the edge line of a skateboard. Then, a second neural network – called the diffusion model – generates the pixels needed to achieve these features.
The diffusion pattern is trained over a series of images in which noise – imperfection – is gradually added to a photograph until it becomes a sea of random pixels. By analyzing these images, the model learns to perform this process in reverse. When you feed it random pixels, it removes noise, turning those pixels into a cohesive image.
At the University of Washington, other university labs and new start-ups, researchers are using similar techniques in their effort to create new proteins.
Proteins start out as chains of chemical compounds, which then twist and bend into three-dimensional shapes that define their behavior. In recent years, artificial intelligence labs like DeepMind, owned by Google’s parent company Alphabet, have shown that neural networks can accurately guess the three-dimensional shape of any protein in the body by based solely on the smaller compounds it contains – a huge scientific breakthrough.