Adobe Firefly

The features that has been proved to be useful and efficient are- 1. Generative Expand- This feature helps me alot when I need to change the size of the image as per the design requirements. 2. Generative Fill- As a designer, many a times you have to remove certain parts of the image before inculcating in the workflow, and it deliver outstanding results 3. Text to Image- This feature is truly useful if we cant find the kind of image we want anywhere. It generates the images many a times true to the description and thus saving a lot of hassle for going for a shoot.

Firefly generates decent content in terms of creativity and uniqueness. I can use the content with a few modifications in our marketing/branding strategy.

The more you use the set of tools at your disposition, the better you'll get at it, and the results will be jaw dropping.

Major weaknesses I encountered in Firefly for images during my workflow are: 1. Inconsistencies with facial features : The AI's results sometimes aren't upto the mark in terms of generating the facial features of people. 2. Feature Error : The reference feature doesn't work well, as it makes the exact replica of reference uploaded, even if the strength is kept minimal. 3. Doesn't provide realistic results : It sometimes doesn't work well if the real life scenarios need to be generated, though give great results with utopian world scenarios.

I have not seen much progress within the last few months in terms of quality of generated content. Other Generative AI products seem to be progressing much quicker. Sometimes it takes quite a few attempts for Firefly to produce the content according to my prompts which can sometimes be very specific.

As with every AI tools out there, rendering/generation can be long, but it's worth the wait, mosf of the time.