How does the Dall-E Mini AI image generator work and how to use it?
The Dall-E Mini uses AI (artificial intelligence) technology to generate photos using a user description of what they are looking for. What do we know so far?
Imagine that instead of searching for an image with specific characteristics like “woman with ice cream cone,” an artificial intelligence image generator created a unique picture based on what you are looking for. This is what OpenAI has been working on.
“DALL·E is a 12-billion parameter version of GPT-3 trained to generate images from text descriptions, using a dataset of text–image pairs,” said the organization.
After releasing DALL-E in early 2021, the second version is now here.
What can you create using DALL-E?
OpenAI has stated that users can create “anthropomorphized versions of animals and objects, combining unrelated concepts in plausible ways, rendering text, and applying transformations to existing images.”
The images that AI has created are quite impressive, particularly those that place text in images that would take serious skills in photoshop to obtain. Additionally, DALL-E has features that would allow a user to make photoshop-like manipulations to an image only through describing, using text, the changes that they would like to see. This could be a game changer in the field of graphic design and while it is not able to do all the work that designers do, it may help reduce the time it takes for them to complete certain edits and adjustments.
The capabilities DALL-E has been designed with are very advanced and carefully crafted to ensure users have a positive experience. For example, if one types in “a collection of glasses sitting on a table,” the AI will produce images with eye glasses as well as cups of different styles.
A more complicated example touted by Open AI is the creation of an image with many small details. For instance, imagine a user is interested in an animation of a “baby penguin wearing blue hat, red gloves, green shirt, and yellow pants.” The software developers noted that while not all the images contained all of the specifications, many did and most followed as least one or two.
The AI also allows users to play around with the textures and style of the image. One could select a photograph-like image while others could chose or animated options. Often, various images with different textures will be created giving the user more options in their selection process.
What is Open AI?
The company who developed DALL-E, Open AI, is based in the Bay Area and seeks “to ensure that artificial general intelligence (AGI)—by which we mean highly autonomous systems that outperform humans at most economically valuable work—benefits all of humanity.”
To fulfill their mission there are limits to the types of images that DALL-E is programed to create. Open AI has stated that violent, hate, and adult images are not able to created using the AI. Efforts have also been made to reduce the risk of photos using the faces of real people, particularly celebrities, from being generated.
Some of Open AI’s top financial supporters include Microsoft, Reid Hoffman’s charitable foundation, and Khosla Ventures.