Although it seems that Artificial Intelligence is something new, in reality many years with us, however, it was not until 2022, when they began to announce a large number of projects that allow them to make the most of this technology for, based on a brief description, automatically generate an image or text. Midjourney, DALL-E, Stable Diffusion, ChatGPT and Bing Chat are the most well-known Artificial Intelligences, but they are not the only ones.
Microsoft announced a few days ago integration of GPT4 in all the applications that form part of Office, an integration that will allow, for example, to generate presentations based on the data of a document, analyze a spreadsheet to summarize the most important points, create a summary of several text documents, generate automatic responses to electronic mails and much more.
How does Firefly work?
I needed to know what Adobe’s plans were to integrate Artificial Intelligence into its applications, plans that have already been unveiled, starting with its name: Firefly. Firefly works exactly the same as the rest of IAs that allow generating images based on a description, but it adds an interesting function not available in it that allows you to establish the range of colors that we want to use, from the type of lighting and from the style of a a much simpler form than other Artificial Intelligences available today.
In addition, we can generate images based on others that we have, that is, using them as a basis to ask Firefly to add modifications in the image. But, the best of all, is that, as we can see in the video of the announcement, all the elements that are generated in the image will be distributed in independent layers, so that we can customize their settings, including the size to adapt it to What we were looking for, apart from moving it to another location without the proposal for the IA, we didn’t just like it.
It also allows us use various images for Firefly to combine them and show the best of each, an option that at the moment is not available in any other IAS that allows us to generate images based on others or based solely on a description. Another interesting function that we offer is the possibility to reduce the grain of the images with low resolution that we want to enlarge. If it is true that this is not a new function, it is to be appreciated that it is also included in Firefly.
Adobe’s AI, they say, has been trained with millions of copyright-free photographs available in Adobe Stock, which represents a huge relief for designers who don’t want to see their designs accused of using an AI based on copyrighted images. by author
Try Adobe’s IA
If we want to try this new Adobe IA, we only need to have an Adobe account and visit the following link where we must enter the data and sign up to the wish list. At the moment Firefly is in beta phase and is not integrated in any of Adobe’s applications, an integration that will reach Photoshop, Express, Lightroom in the next few months and that will also be available independently through Adobe Creative Cloud, for lo que será necesario passar por caja para poder utilizarla.
