Make A video and DALL-E 2
The advancements in AI have made giant leaps in the past year.
Open AI Dall-e2 is an artificial intelligence that reads written text requests from users and then, through its knowledge of millions of images. creates its own entirely new images pixel by pixel..
The results are incredibly good, which is why since its limited release in April of this year, Dall-e 2 has evoked both fascination and trepidation.Could this new AI replace our graphic designers and illustrators?
Powered by a neural network, a type of algorithm that mimics how the human brain works, this AI is designed to imitate human thinking. It learns what objects are and how they relate to each other by analyzing images and the captions that accompany them.
It is akin to introducing a child to flash cards. If there are a lot of pictures of a bear captioned ‘bear’ the AI will learn what a bear looks like.
If you type in ‘a bear with an umbrella in the rain’ it draws on what it knows about bears, umbrellas and rain to put together a logical interpretation.
Drawing on what programs like Dalle have achieved, a new team is developing MAKE-A-VIDEO which generates high quality video clips from text prompts by teaching AI what the world looks like by combining already existing text to image data with motion. It does this by collecting large amounts of unsupervised video footage as data. All the data used are publicly available datasets from the internet and the results, as evidenced by the examples provided, are quite astounding. We cannot wait to see what this exciting project will enable creators and artists to do in the near future.
