NÜWA is a unified multimodal pre-trained model that can generate new…
Or, moreover manipulate existing visual data.
Furthermore, imagine for example images and videos being manipulated to the most perfecet degree and the fooling that could go on across the world!
Moreover NÜWA or 8 visual synthesis tasks: https://github.com/microsoft/NUW
In addition, it achieves SotA results on text-to-image generation, text-to-video generation, video prediction, etc. Outperforms DALL-E in text2image.
“Do not look at where we are. Look at where we will be two more papers down the line.”
— Károly Zsolnai-Fehér
What will this model mean for data and the ability to manipulate it?
As a result of the availability of easy to access models that can help nefarious actors or nation states create fraud, this new model seems to be quite frightening. What is the positive that this model offers for society and community as a whole? A pre-trained model used by anyone around the world, how will that help students or teachers educated the next generation? Furthermore, how will this help feed children across the globe? At least Bitcoin offers value for poor people with no currency to rely on to have a vehicle to deposit their monetary value in without having to worry about waking up the next morning with a bank account vanished.
In conclusion, this might be helpful for certain academic researchers or maybe police organizations across the world establish what is true vs a deep fake? Lastly, what will society’s gain be from this new model?