DALL-E, a text-to-image program, is taking the internet by storm as it converts words into images through artificial intelligence models. DALL-E is described as a new AI system that can create realistic images and art from a description in natural language. If you are wondering about the name, well DALL-E is a blend of the names Salvador Dali and Wall-E.
This program is just one of several new AI image tools that offer revolutionary promise. OpenAI’s DALL-E, first introduced in January 2021, produces a wide range of images in seconds. The new Program is super accessible and can be accessed from any DALL E website. The updated program DALL-E 2 has been built to revolutionize the way people use AI with images. Basically, it’s a powerful new iteration that performs at a much higher level.
A huge improvement has been made to the image resolution, lower latencies – meaning how long the image takes to be created – and a more intelligent algorithm for creating images. The program amplifies what a person can do but is dependent on the creativity of the person using it. The user can add different art techniques as per request, inputting styles of drawing, oil painting, knitted out of wool, plasticine model, or a cave painting/drawing, etc.
Aditya Ramesh, one of the lead engineers on DALL-E 2, highlighted two clever techniques – painting and variations – which work in a similar way to the rest of DALL-E, but with a twist. A person using the inpainting technique can take an existing image and edit new features into it or change parts of it. While with the variations technique, the person has to feed in an image or illustration, and DALL-E’s variation tool will make hundreds of its own versions. The tool can also be used to combine two images into one.
Joanne Jang, an engineer with Open AI, said they are working hard to accelerate access, but it will take some time. He revealed that they had invited 10,217 people, as of June 15, to try DALL-E. The system that is publicly available is DALL-E Mini. It uses open-source code from a loosely organized team of developers.
Critics have highlighted the potential for harm that could come from a model like DALL-E 2 that has learned to make images by essentially scouring the internet. To address this, developers with OpenAI removed violent content from training data. They placed filters to stop DALL-E 2 from generating images if users submit prompts that may violate company policy.
OpenAI is excited about the opportunities that the technology can enable. It has the potential to open up creative opportunities for individuals and help with commercial applications for interior design or websites.