gizmodo.com: On Tuesday, Adobe introduced its own generative AI model family called Firefly, claiming—like young people tend to do in the throes of pubescence—it’s not like the other scrappy AI image generators. Instead of using training data dredged up from the internet, Adobe said its crafted its service only using images on its stock image site, openly licensed work, and public domain content.
4 comments:
I still don’t know how to feel about a large company like Adobe providing AI programs. I’ve been very worried about AI using “stolen” images or images that just belong to other people With Adobe saying they are only using “stock imaging sites, openly licensed work, and public domain content” to make the base for their program, I don’t really know how to feel. I guess knowing that this program is using these sources makes me feel a bit better. I don’t like AI when it takes literally every image that it can to form its base. I also think AI is a cheap and easy way out of being creative. But if a computer and program doesn’t have as many sources, it makes it a bit better. It’s still weird that a computer can create art and this is the world we’re going to have to live in now.
If I’m going to be honest, I’m just not sure if I find this to be realistic or if I believe it. Maybe this is due to the fact that I have read so many articles posted on here about AI images being “stolen” or about how people’s art is being used on AI… So maybe it is just a bias, but I guess I’m not sure how such a large brand like adobe, especially a brand with so much to do involving images, is saying that their AI will not “steal” art or pictures from anywhere. I do think that the possibilities of Adobe-powered AI are very strong, but the idea also seems off (to me) due to this article. I’ll be interested to read future articles and see how this technology develops, as well as what the public’s opinion is. It’s interesting to read about the differences between different AI software’s as more brands come out with them.
This is a really interesting topic, since the beginning of AI generating arts, the conversation around copy rights and how AI often uses other people’s work as reference and copy their style and even techniques that artists would using in drawings and their artworks. In this situation and an intense ongoing conversation about the copyright issues around it, the creation of Adobe creating and making an AI that does not use other artists work is ground breaking and scary in what they are capable of doing. With this idea, I’m really interested to learn more about what it has as its training data. This is because since it is not using other artists’ work, does it means that developers within Adobe creates artwork to train it or does it mean that Adobe have generated a machine that understands what is artistic and able to “think” on their own without any pictures.
Personally I am very against AI image generators. A lot of artists that I have followed for years now have seen instances where AI image generators have taken their work and used it in their generated images. The sourcing of images is very shady to me. The fact that there has really not been any actual compensation to anyone whose images are used in these AI image generators does not sit very well with me at all. I also just do not really understand why people feel the need to use AI In the first place. Maybe I just need someone to explain it better why they would need to use AI. The influx over the past couple of years of different AI image generators is also slightly concerning specifically for artists and photographers because it makes it harder and harder with each new one that comes out to make sure that your work is protected from being used without your permission because there are just so many AI image generators.
Post a Comment