Artificial Intelligences “artists” are increasingly common and popular. OpenAI and Google have two of the most relevant examples today, but now they are being joined by Microsoft, a company that has just created its own system to generate images from text, extend the design of an existing work or convert it into video .
is the name of the technology. Microsoft describes it as “a multimodal generative model designed to generate high-quality images and video from text, image, or video input.” Therefore, it is not too far from DALL-E, from OpenAI, or Brain, from Google.
The big difference between NUWA-Infinity and these systems is that it can generate long-duration videos through a description, as well as high-resolution images with arbitrary sizes.
Regarding images, its main feature is that it can “extend” any image. On its page, Microsoft uses Van Gogh’s The Starry Night as an example to show how you can add more elements to the painting without forgetting the original design and giving it a very exact continuity.
Within the sample page there are also other examples, such as the iconic photo of a Windows landscape, in which you imagine what other elements are around, or videos generated from photographs.
Without a doubt, the field of AI-generated images has become one of the most popular uses today, so companies are paying more attention to its development.
About Brain, Google has shared that these are important issues, because beyond creating new images, the idea of these tools is to strengthen artificial intelligence systems from data that make them even more sophisticated and try to reduce prejudice. that these technologies can reproduce.
Audits of data sets have revealed that they tend to reflect social stereotypes, oppressive viewpoints, and derogatory or harmful associations with marginalized identity groups.
On the other hand, DALL-E announced a few days ago that its tool will open a beta version for one million users; however, as soon as the news was announced, the waiting list was filled, demonstrating the interest of users in making their dreams come true in a work.
It is worth mentioning that OpenAI is already monetizing its tool, since those users who get access will be given 50 free credits during their first month. Subsequently, users will have 15 credits each month and if it is not enough they will have to pay $15 for 115 credits.
The company has said it wants to curb misuse of its tool before a broader rollout. For this reason, they are taking measures to block content with realistic faces and thus avoid the creation of “violent, adult or political” content.
However, something relevant about DALL-E is that users will also obtain the rights to use the images, even for commercial uses, so they can be monetized in any way they decide.
