After waiting long and hard, an invitation to create with DALL-E appeared in my email and here is my experience with the model.
DALL-E is a model developed by OpenAI that generates images from natural language descriptions. We’ve been hearing about AI generated art for some time and we’ve seen it’s ability to generate hyper-realistic images. So much so that OpenAI has restricted actual faces of people from appearing in any of the generated art. Imagine an AI generated image of a real world police officer accepting a bribe.
This police officer does not exist, it is a product of an AI model which I asked to generate a realistic image of a police officer taking a bribe. See the problem in the system if I could ask the AI model to use the face of an actual cop?
Now that we’ve seen what it is capable of doing, I’ll share some other artwork that I was able to create using DALL-E.
“schrodinger’s cat doing drugs”
“a man shredding a guitar on top of a fighter jet”
“A snake picking up dumbells. ”
“Two planets colliding in the sky over a city at night”
“a sufi astronaut dancing in space ”
“Lahore during a starry night. ”
“A camel driving a taxi in Lahore. ”
I didn’t quite get the results I was looking for, so let’s try this again.
“A camel driving a taxi in time square”
“a teddy bear who is a barista at a gourmet restaurant ”
“Dall E shaking hands with Van Gogh ”
