I lead a workshop on text to image topic during "la journée des artistes" in France on 10th & 11th of june 2023. Objective was to make the public discover image creation with AI by acting as creators themselves, and with my assistance when needed.
Audience was mainly professional and amateur photographs and visitors interested in photos, paintings and graphic arts. Thanks to Adobe with the offered Beta version, I could demonstrate Firefly to 100's of visitors.
All of the visitors could leave the workshop with their “co-creation” (with the Adobe watermark on it !)
Here bellow is a short feedback on reactions:
I tested a number of “text to image” application and I must admit that Adobe firefly (Web Beta version) was the most ergonomic and easiest to use for me and was the perfect tool for my public workshop.
This being said, I found the outputs to be of slightly lower quality, especially when looking for ultra realistic outputs compared to a solution like Midjourney.
Limitations, which do not exist in other application, regarding certain topics and the exclusion of proper nouns (people, brand, monuments, etc.) was a bit annoying but not an obstacle considering we were in a Beta version.
Photographs from older generation (say 60+) where somehow interested and impressed by the results and the potential, but they were also a bit skeptical and even scared about the AI technology wave that could jeopardize their jobs. Copyright and sources of images were also a concern.
Photographs from younger generation who are already well involved in digital things were very active and practical to create some outputs they could use immediately (noncommercial use) and were generally very impressed with first shots, with no need for numerous iterations.
Surprisingly (to me) the most demanding public where young kids (as young as 6 to 10 years old), asking for very imaginative and precise outputs and it was eventually hard to meet their expectations.
I will finish with an anecdote in the poetry field, one amateur was looking for an illustration of her poem. She had no special expectation, but the first shot made her very impressed and tears of emotion can be read on her face, which in a certain sense illustrated how powerful text to image can be …
I would like to thank Adobe for having given me the opportunity to test Firefly !
Note : attached is the screen saver poster (in french) i created with Firefly assifor the workshop