MicrostockGroup
Agency Based Discussion => Adobe Stock => Topic started by: MatHayward on March 21, 2023, 08:05
-
Hi Everyone,
This morning Pacific Daylight Time, we emailed our contributors the following message.
“We wanted you to be the first to know that today Adobe introduced Adobe Firefly, a new family of creative generative AI models. Firefly’s initial focus on images and test effects is now available in beta at firefly.adobe.com. The first model of Firefly is trained on Adobe Stock images, openly licensed content, and public domain content, where copyright has expired and is designed to generate images safe for commercial use. During the beta phase, the Firefly-generated assets cannot be used for commercial purposes.
We are investing in generative AI to help creators transform the way they imagine and explore possibilities. We are committed to developing generative AI responsibly, with creators at the center. Through efforts like the Content Authenticity Initiative, we’re standing up for accountability, responsibility, and transparency in generative AI. We are developing a compensation model for Stock contributors, and we will share the details of this model when Firefly exits beta.”
You can read more about our announcement here: https://blog.adobe.com/en/publish/2023/03/21/bringing-gen-ai-to-creative-cloud-adobe-firefly
Please see our FAQ for more information about the beta release of Firefly here: https://helpx.adobe.com/stock/contributor/help/firefly-faq-for-adobe-stock-contributors.html
Learn more about the Content Authenticity Initiative here: https://contentauthenticity.org/
Please let me know if you have any questions,
Mat Hayward
-
Hi Everyone,
This morning Pacific Daylight Time, we emailed our contributors the following message.
“We wanted you to be the first to know that today Adobe introduced Adobe Firefly, a new family of creative generative AI models. Firefly’s initial focus on images and test effects is now available in beta at firefly.adobe.com. The first model of Firefly is trained on Adobe Stock images, openly licensed content, and public domain content, where copyright has expired and is designed to generate images safe for commercial use. During the beta phase, the Firefly-generated assets cannot be used for commercial purposes.
We are investing in generative AI to help creators transform the way they imagine and explore possibilities. We are committed to developing generative AI responsibly, with creators at the center. Through efforts like the Content Authenticity Initiative, we’re standing up for accountability, responsibility, and transparency in generative AI. We are developing a compensation model for Stock contributors, and we will share the details of this model when Firefly exits beta.”
You can read more about our announcement here: https://blog.adobe.com/en/publish/2023/03/21/bringing-gen-ai-to-creative-cloud-adobe-firefly
Please see our FAQ for more information about the beta release of Firefly here: https://helpx.adobe.com/stock/contributor/help/firefly-faq-for-adobe-stock-contributors.html
Learn more about the Content Authenticity Initiative here: https://contentauthenticity.org/
Please let me know if you have any questions,
Mat Hayward
This 👇 is not good. :(
Can you please explore the opt-out faster?
-
That is amazing Mat, looking forward to get access to the beta program to try it out.
-
We are developing a compensation model for Stock contributors, and we will share the details of this model when Firefly exits beta.
Universal Basic Income? ;)
Re-establish € payment for europeans? ???
-
Thank you!!
Can‘t wait to try it.
Everything generated with firefly can be used safely and uploaded as ai content. That is fantastic!!
ETA:
I just applied to the program, hope I get accepted for both images and text.
-
“ Everything generated with firefly can be used safely and uploaded as ai content. That is fantastic!!”
But not during beta.
Can’t wait to hear the compensation model. UBI is a good way to look at it. Technology that displaces workers/creatives should help to fund those displaced so they can continue to provide alternatives.
-
True, we have to wait for beta to end. But perhaps Adobe can find a way to give earlier access for stock producers.
-
True, we have to wait for beta to end. But perhaps Adobe can find a way to give earlier access for stock producers.
Already done, microprocessor have access! since they are the new stock producers ;D
-
LOL! I am still hoping to have a few years of fun with ai, before I am completely replaced…
-
I always thought that company must "explore all the possibilities" of such a complex product before development even starts, especially considering that contributor images are the key core asset of this product creation.
And the only reason not to give an opt-out (i'm not even talking about "opt-in" lol), is because, you know... because... mmm... But what do i know, right? ;D
-
I was looking at coverage of Adobe's Firefly announcement, coming as it does on a day when Google has announced access to Bard - following lots of coverage of lots of chatbots. The most fun headline was from a Forrester article: "Generative AI Gets An Upgrade To Business Class With Adobe Firefly"
https://www.forrester.com/blogs/generative-ai-adobe-firefly/ (https://www.forrester.com/blogs/generative-ai-adobe-firefly/)
Gizmodo says "Adobe Says Its New ‘Firefly’ AI Image Generator Doesn't Steal Other People's Art" and at the end points out: "However, in Adobe’s case, it seems stock contributors won’t have much or any choice at all. " It was the only article I found that talked about Shutterstock's launch of something similar while Getty went the route of suing Stable Diffusion:
https://gizmodo.com/adobe-ai-ai-art-generator-dall-e-firefly-1850247670 (https://gizmodo.com/adobe-ai-ai-art-generator-dall-e-firefly-1850247670)
Tech Crunch has a long article about the launch, competitors, and the issues yet to be worked out:
https://techcrunch.com/2023/03/21/adobe-firefly-generative-ai/ (https://techcrunch.com/2023/03/21/adobe-firefly-generative-ai/)
The Verge talks about this being a big deal as Adobe sits at the center of the creative ecosystem:
"Adobe is putting one big twist on its generative AI tools: it’s one of the few companies willing to discuss what data its models are trained on. And according to Adobe, everything fed to its models is either out of copyright, licensed for training, or in the Adobe Stock library, which Costin says the company has the rights to use. That’s supposed to give Adobe’s system the advantages of not pissing off artists and making its system more brand-safe. “We can generate high quality content and not random brands’ and others’ IP because our model has never seen that brand content or trademark,” Costin said."
https://www.theverge.com/2023/3/21/23648315/adobe-firefly-ai-image-generator-announced (https://www.theverge.com/2023/3/21/23648315/adobe-firefly-ai-image-generator-announced)
The Guardian's tech columnist covers the current chatbot landscape and points out the importance of Adobe being upfront about what Firefly was trained on:
"A core plank of Firefly is that the company is offering “safe” generation: its generative model is, it says, “trained on Adobe Stock images, openly licensed content, and public domain content where copyright has expired”. In other words, if you work with Firefly-created images, you know for certain that there is no nasty copyright lawsuit coming down the line. That stands in stark contrast to GPT-4, which is trained on … well, no one actually knows."
https://www.theguardian.com/technology/2023/mar/21/the-ai-tools-that-will-write-our-emails-attend-our-meetings-and-change-our-lives (https://www.theguardian.com/technology/2023/mar/21/the-ai-tools-that-will-write-our-emails-attend-our-meetings-and-change-our-lives)
C-NET (who had to put a note at the end about how it uses AI to write articles, although apparently not this one) covered the legal as well as technical aspects of how what Adobe is doing differs from others venturing into the generative AI world:
"Developing good AI isn't just a technical matter. Adobe set up Firefly to sidestep legal and social problems that AI poses."
https://www.cnet.com/tech/computing/adobe-our-new-generative-ai-will-help-creative-pros-not-hurt-them/ (https://www.cnet.com/tech/computing/adobe-our-new-generative-ai-will-help-creative-pros-not-hurt-them/)
And a few more..
https://www.axios.com/2023/03/21/adobe-generative-ai-tools-firefly (https://www.axios.com/2023/03/21/adobe-generative-ai-tools-firefly)
https://appleinsider.com/articles/23/03/21/new-adobe-firefly-will-use-ai-to-generate-art-but-protect-artists (https://appleinsider.com/articles/23/03/21/new-adobe-firefly-will-use-ai-to-generate-art-but-protect-artists)
https://www.cnbc.com/2023/03/21/adobe-firefly-generative-ai-lets-you-type-to-edit-images.html (https://www.cnbc.com/2023/03/21/adobe-firefly-generative-ai-lets-you-type-to-edit-images.html)
https://www.fastcompany.com/90868402/adobe-firefly-generative-ai-photoshop-express-illustrator (https://www.fastcompany.com/90868402/adobe-firefly-generative-ai-photoshop-express-illustrator)
https://www.techradar.com/news/nvidia-launches-new-ai-service-and-adobe-is-on-board (https://www.techradar.com/news/nvidia-launches-new-ai-service-and-adobe-is-on-board)
Looking at what the stock market thinks, Adobe and Google's stock is up today (more than the overall market), Microsoft's is down slightly and Shutterstock's was down but is now neutral.
What's that quote about a curse? May you live in interesting times? :)
-
Those are great links, I'll add them to my other list/thread.
Thank you!
-
for a more general discussion including AI replacing programmers, editors & managers ("anything that can be done remotely on a computer)
A.I. Is About to Get Much Weirder. Here’s What to Watch For.
The Vox writer Kelsey Piper talks about the increasing pace of A.I. development, how it’s changing the world and what to do about it.
https://www.nytimes.com/2023/03/21/opinion/ezra-klein-podcast-kelsey-piper.html (https://www.nytimes.com/2023/03/21/opinion/ezra-klein-podcast-kelsey-piper.html)
March 21, 2023
-
Adobe Firefly, I did not understand anything. Will I be able to create images in Adobe Firefly in the future and then sell on Adobe?
-
Here people are already creating videos using AI.
https://www.bloomberg.com/news/articles/2023-03-20/generative-ai-s-next-frontier-is-video (https://www.bloomberg.com/news/articles/2023-03-20/generative-ai-s-next-frontier-is-video)
-
Adobe Firefly, I did not understand anything. Will I be able to create images in Adobe Firefly in the future and then sell on Adobe?
During the Firefly beta, any Firefly-generated assets cannot be used for commercial purposes. We will share details about the use of Firefly-generated assets once Firefly is officially released. Adobe Stock today accepts AI generated works that otherwise meet Adobe Stock’s requirements.
Thanks for the question,
Mat Hayward
-
Mat, what if I use AI to transform my original image?
Will it be considered as AI or original image?
-
Mat, what if I use AI to transform my original image?
Will it be considered as AI or original image?
I'm assuming you mean this in the context of uploading to Adobe Stock? If you are using any generative AI software during any part of the creation of your asset, you must comply with all generative AI submissions. As noted above in this thread, while Firefly is in beta, you cannot use the content generated for commercial use, which includes stock uploads.
As a reminder, the generative AI submission guidelines can be found here: https://helpx.adobe.com/stock/contributor/help/generative-ai-content.html
Thanks,
Mat Hayward
-
Hi Everyone,
This morning Pacific Daylight Time, we emailed our contributors the following message.
“We wanted you to be the first to know that today Adobe introduced Adobe Firefly, a new family of creative generative AI models. Firefly’s initial focus on images and test effects is now available in beta at firefly.adobe.com. The first model of Firefly is trained on Adobe Stock images, openly licensed content, and public domain content, where copyright has expired and is designed to generate images safe for commercial use. During the beta phase, the Firefly-generated assets cannot be used for commercial purposes.
We are investing in generative AI to help creators transform the way they imagine and explore possibilities. We are committed to developing generative AI responsibly, with creators at the center. Through efforts like the Content Authenticity Initiative, we’re standing up for accountability, responsibility, and transparency in generative AI. We are developing a compensation model for Stock contributors, and we will share the details of this model when Firefly exits beta.”
You can read more about our announcement here: https://blog.adobe.com/en/publish/2023/03/21/bringing-gen-ai-to-creative-cloud-adobe-firefly
Please see our FAQ for more information about the beta release of Firefly here: https://helpx.adobe.com/stock/contributor/help/firefly-faq-for-adobe-stock-contributors.html
Learn more about the Content Authenticity Initiative here: https://contentauthenticity.org/
Please let me know if you have any questions,
Mat Hayward
Just wondering how you can keep a straight face announcing this exciting news. Pay must be really good then at Adobe :)
Are they looking for more people advocating this 'without you there is no us' stuff? Maybe some (other) people would change jobs then. Asking for a friend.
-
What about the 2.6 millions AI generated images already uploaded and actually for sale in Adobe Stock Photo web site? Most of them generated by Midjourney
-
We are investing in generative AI to help creators transform the way they imagine and explore possibilities. We are committed to developing generative AI responsibly, with creators at the center.
Mat Hayward
Calling Ai users "CREATORS" is like to call people with a tv remote control - MOVIE DIRECTORS.
-
It doesnt look that ordinary customer will be able to use firefly such easy like to search image in stocks. It looks like you will need some knowledge how to use this app, it is some kind advanced photoshop or illustrator. We need to adapt to the changes, there is no time to cry like "They are stealing our jobs".
-
It doesnt look that ordinary customer will be able to use firefly such easy like to search image in stocks. It looks like you will need some knowledge how to use this app, it is some kind advanced photoshop or illustrator. We need to adapt to the changes, there is no time to cry like "They are stealing our jobs".
Welcome newcomer, you are late you know? you should have bebun stock photography 15 years ago...
You use a smartphone to produce your images, and already complained about sales on Adobe Stock : "[...] My sales dropped drastically a couple of days ago". So, you do know what crying is ;D See you soon for your next complain.
-
It doesnt look that ordinary customer will be able to use firefly such easy like to search image in stocks. It looks like you will need some knowledge how to use this app, it is some kind advanced photoshop or illustrator. We need to adapt to the changes, there is no time to cry like "They are stealing our jobs".
Welcome newcomer, you are late you know? you should have bebun stock photography 15 years ago...
You use a smartphone to produce your images, and already complained about sales on Adobe Stock : "[...] My sales dropped drastically a couple of days ago". So, you do know what crying is ;D See you soon for your next complain.
oh gosh, your life must be really meaningless if you look at my post history and try guess my life. poor you :D
-
Hi Everyone,
This morning Pacific Daylight Time, we emailed our contributors the following message.
Please let me know if you have any questions,
Mat Hayward
I have this main question: I didn't receive anything at all ... Are there any settings (which ones?) that I have failed to set correctly? Can you help me? I try not to get spammed too much in general (not by Adobe, in general!) , but real communications I always want to have... Maybe in the year of transition from Fotolia I missed something and can't find it anymore, is that possible? Can you help? :)
Or maybe not all contributors were involved, is that possible?
-
Will you compensate contributors if their content was used to train Adobe Firefly to generate outputs?
We are developing a compensation model for Stock contributors, and we will share the details of this model when Firefly exits beta.
so in other words in the beta phase adobe use our content for free ??? typical pr speak. just don't name the problem.
-
Will you compensate contributors if their content was used to train Adobe Firefly to generate outputs?
We are developing a compensation model for Stock contributors, and we will share the details of this model when Firefly exits beta.
so in other words in the beta phase adobe use our content for free ??? typical pr speak. just don't name the problem.
Yeah, and it's not like once our images have been used to train the AI in the beta phase, our images will still be needed after that, so there won't be anything to "compensate" us for anymore once beta is over. ::)
-
[...] oh gosh, your life must be really meaningless[...] poor you :D
;D imagine my life the way you want, judge me if you want, it makes me smile.
I am sorry to have hurt you that much.
-
Zorba, you need to be opted in to receive marketing emails in order to receive Adobe Stock Contributor related emails. Send me a private message here with your Adobe ID email address and I can check if we have the right email on your contributor account.
-
Zorba, you need to be opted in to receive marketing emails in order to receive Adobe Stock Contributor related emails. Send me a private message here with your Adobe ID email address and I can check if we have the right email on your contributor account.
Thanks Kirsten for letting us know, after the fact, that our photos were used to train Adobes new AI tool. Can you eloborate a bit about your compensation plan for this?
It is not that we have had a chance to opt out or is it somewhere in the fine print?
-
Zorba, you need to be opted in to receive marketing emails in order to receive Adobe Stock Contributor related emails. Send me a private message here with your Adobe ID email address and I can check if we have the right email on your contributor account.
Thanks Kirsten for letting us know, after the fact, that our photos were used to train Adobes new AI tool. Can you eloborate a bit about your compensation plan for this?
It is not that we have had a chance to opt out or is it somewhere in the fine print?
I think you've misunderstood, Kirsten is talking about the missing announcement mail, nothing about photos included.
By the way I've missed too, so I'm sending a pvt message :)
-
Zorba, you need to be opted in to receive marketing emails in order to receive Adobe Stock Contributor related emails. Send me a private message here with your Adobe ID email address and I can check if we have the right email on your contributor account.
I sent you some other private messages but I've never seen answers... the last was about PNG :)
But thanks I'll check by myself about that! thanks! I want to keep this profile very private and non-retrievable :)
Adobe is ok, but there are some other companies that are not so fair ! :-D
-
Zorba, you need to be opted in to receive marketing emails in order to receive Adobe Stock Contributor related emails. Send me a private message here with your Adobe ID email address and I can check if we have the right email on your contributor account.
Thanks Kirsten for letting us know, after the fact, that our photos were used to train Adobes new AI tool. Can you eloborate a bit about your compensation plan for this?
It is not that we have had a chance to opt out or is it somewhere in the fine print?
I think you've misunderstood, Kirsten is talking about the missing announcement mail, nothing about photos included.
By the way I've missed too, so I'm sending a pvt message :)
She's speaking for Adobe and I know she reacted to something else, I can read, just wanted to ask her a question :)
-
To clarify, Kirsten is referencing the question from people who may not have received the actual email yesterday. As she stated, you must be opted in to receive marketing emails via adobe.com to receive this sort of information in real time.
You can be assured that any major announcements like the news about Firefly will be called out here in Microstock Group Forum too. :)
Thank you,
Mat Hayward
-
To clarify, Kirsten is referencing the question from people who may not have received the actual email yesterday. As she stated, you must be opted in to receive marketing emails via adobe.com to receive this sort of information in real time.
You can be assured that any major announcements like the news about Firefly will be called out here in Microstock Group Forum too. :)
Thank you,
Mat Hayward
Getty images had their images pilfered to train an AI and they are in the process of mounting a significant legal action against the company that did so.
Shutterstock compensate it's contributers for images that have been used in the data sets to train their A.I. you are also not able to opt out.
Adobe is compensating their contributers how. I think that ignoring this question which has already been asked doesn't make for a good look. So what's the deal. This isn't light and fluffy ... awwww cummon guys ... this is business. You don't get free use. It is a requirement to be transparent regarding payment for use. what Matt.
-
@MatHayward
Is there any chance that there will be Ai search term in filters or category in search panel on site to help users / buyers decide what kind of artwork or "airtwork" they wish ?
Or at least something like a subcategory for people - people with 6 fingers, people with 7 fingers ...you know....something like that?
-
@MatHayward
Is there any chance that there will be Ai search term in filters or category in search panel on site to help users / buyers decide what kind of artwork or "airtwork" they wish ?
Or at least something like a subcategory for people - people with 6 fingers, people with 7 fingers ...you know....something like that?
;D
They need an extra finger. AI models have the right number of fingers, once the supernumerary ends up somewhere inside traditional photographers... ::)
-
Saw the Adobe CEO on some business channel discussing this. He was saying they are exploring ways for artists to opt out of AI training or be compensated etc. I was just thinking it sounded quite positive, then he said something dismissively like “it’s initially just using public domain and our stock library anyway”. Realized he didn’t seem mean us when he said “artists”?
-
Saw the Adobe CEO on some business channel discussing this. He was saying they are exploring ways for artists to opt out of AI training or be compensated etc. I was just thinking it sounded quite positive, then he said something dismissively like “it’s initially just using public domain and our stock library anyway”. Realized he didn’t seem mean us when he said “artists”?
We - contributors - are not of much concern to the CEO of a $16+ billion company the bulk of whose income comes from subscriptions. Keeping the subscriptions going and growing is what matters to him. We need to be managed, but IMO we aren't of any concern unless the flow of subscriptions is at risk.
The Do Not Train tag is a promise, not something available today. I'm guessing that any use of it in material submitted to Adobe Stock would be disallowed, like putting any visible watermark on your uploads is today. In other words that wouldn't give contributors to Adobe Stock an opt out.
I had never heard of Plagiarism Today until I did a search after reading your post, but they have some interesting observations about the legal wrangles ahead.
https://www.plagiarismtoday.com/2023/03/22/can-adobe-and-nvidia-fix-ais-copyright-woes/ (https://www.plagiarismtoday.com/2023/03/22/can-adobe-and-nvidia-fix-ais-copyright-woes/)
"Even if artists and photographers did sign away these rights in their contracts, that doesn’t mean that they approve of this use or direction. Though they might not have legal recourse, that doesn’t mean they won’t consider removing their work from the library or stop working with them for future work.
The agreement to pay royalties can, at least theoretically, go a long way to soothing potential tensions in this space. But the details are both important and unknown. We don’t know how Adobe or Nvidia will pay royalties, and we don’t know how much those royalties will be worth.
This is going to be a space to watch moving forward, whether the photographers and artists in those libraries feel that they are getting a fair deal."
Reuters says that NVIDIA trained their system on images licensed from Getty, SS and Adobe - perhaps that's in error. How would it be OK for Adobe to let a third party use Adobe Stock's library for training - that's not covered by the contributor terms of service:
https://www.reuters.com/technology/adobe-nvidia-ai-imagery-systems-aim-resolve-copyright-questions-2023-03-21/ (https://www.reuters.com/technology/adobe-nvidia-ai-imagery-systems-aim-resolve-copyright-questions-2023-03-21/)
Edited to add that NVIDIA's press release (look at the bottom) describes their partnerships with Adobe, Getty and Shutterstock. I think it says that Adobe will be using NVIDIA's services to create an Adobe product, but there's some mention of an NVIDIA product, Picasso, that will also use what's jointly developed.
https://nvidianews.nvidia.com/news/nvidia-brings-generative-ai-to-worlds-enterprises-with-cloud-services-for-creating-large-language-and-visual-models (https://nvidianews.nvidia.com/news/nvidia-brings-generative-ai-to-worlds-enterprises-with-cloud-services-for-creating-large-language-and-visual-models)
https://nvidianews.nvidia.com/news/adobe-and-nvidia-partner-to-unlock-the-power-of-generative-ai (https://nvidianews.nvidia.com/news/adobe-and-nvidia-partner-to-unlock-the-power-of-generative-ai)
Sounds as though this is skating right on the line between what's permitted in our contributor agreement (developing new Adobe products) and giving away our intellectual property (NVIDIA getting a new product powered by Adobe Stock contributors' content).
Ars Technica ends its article (https://arstechnica.com/information-technology/2023/03/ethical-ai-art-generation-adobe-firefly-may-be-the-answer/amp/) saying "As always, we'll need to take Adobe's claims with a grain of salt, and we'll keep you updated as new details emerge."
This is an investor and business focused take on the Firefly announcement, where the blurring of the meaning of creators comes up again - does it mean us, the source material, or those using Adobe tools built in part from our sources?
https://www.techtarget.com/searchcustomerexperience/news/365533294/Adobe-Firefly-brings-generative-AI-imaging-marketing-tool (https://www.techtarget.com/searchcustomerexperience/news/365533294/Adobe-Firefly-brings-generative-AI-imaging-marketing-tool)
"Don Fluckinger, an analyst at TechTarget's Enterprise Strategy Group, added that Adobe's proposed safeguards for preserving creators' legitimacy is important for the future of generative AI.
"Adobe senior leadership's thoughtful approach to 'GenAI' with an eye toward protecting creators -- and making their creations commercially safe -- looks like a huge step toward legitimizing the technology for business use," Fluckinger said."
-
Great summary. Based on that interview I very much got the impression that the CEO considered images in the AdobeStock library as their property to do with as the wish. All his concern (genuine or not) about artists’ rights very much seemed to be with regards to work outside of AS's library. It's pretty crazy when you think how for out of the scope of any original agreement this sort of usage is.
-
@MatHayward
Is there any chance that there will be Ai search term in filters or category in search panel on site to help users / buyers decide what kind of artwork or "airtwork" they wish ?
Or at least something like a subcategory for people - people with 6 fingers, people with 7 fingers ...you know....something like that?
already there just include -ai in your search, since AS requires that keyword
-
Great summary. Based on that interview I very much got the impression that the CEO considered images in the AdobeStock library as their property to do with as the wish. All his concern (genuine or not) about artists’ rights very much seemed to be with regards to work outside of AS's library. It's pretty crazy when you think how for out of the scope of any original agreement this sort of usage is.
True,
everything is after the fact. We had never any choice. And they are still thinking if they will or will not compensate us. But you never had a choice to opt out to start with because they allready did it, using our photos to train their AI.
-
And specifically to @MatHayward ,
Adobe had the possibility to ask permission, upfront, from contributors when they started this firefly project. But they apparently didn't. What is the corporate response to this? Adobe must have known that, this project, was pretty sensitive seeing issues across the industry. Or was this project started already before issues arised? If you want to be open to contributors, as you always pretend to do, then this is the time to do so.
-
@MatHayward
Is there any chance that there will be Ai search term in filters or category in search panel on site to help users / buyers decide what kind of artwork or "airtwork" they wish ?
Or at least something like a subcategory for people - people with 6 fingers, people with 7 fingers ...you know....something like that?
;D
They need an extra finger. AI models have the right number of fingers, once the supernumerary ends up somewhere inside traditional photographers...
There is an expression in my country :
" What grandma speaks about, that is what grandma likes."
So please stop this joker wannabee attempts and keep those fingers somewhere inside yourself.
-
I think there is missunderstanding here about cynical and second degree.
It seems to be you are insulting, I never had such intention. I was with you in fact and have cliked +1 on your message before writing. So thx to reconsider your deductions. I wrote AGAINST those A(supposed I) tools.
-
I don't understand why contributors have not been a) notified earlier of the fact that Firefly is in beta and has been learning from Adobe Stock's library, i.e. the assets uploaded by contributors, and b) compensated already for use of their assets in Firefly's training.
Why is Adobe only at the 'exploring' stage of an opt-out possibility and still busy 'developing' a compensation model? Shouldn't that be the first thing on the priority list?
Once again it looks like contributors and compensation of them are treated as an afterthought, a nuisance, an annoying expense.
-
Now they will use our work and pay us only as much as they themselves decide to pay us. In the next phase, which is already underway, AI users will provide more subtle feedback and train Ai to the point where she can train herself again on selection of Ai generated images. Also, the user interface will be simplified for use by non-artists and the worst flaws (like fingers or strange faces) will be filtered out. At that point, no one will need us anymore. After all, they won't pay us compensation forever.
In the worst case, they will do without new images from us. They have huge library already. Mass action by contributors is unlikely. They will just pay us something (they decide how much) to make it legal.
So our value is small and temporary. They just don't care.
I don't understand why contributors have not been a) notified earlier of the fact that Firefly is in beta and has been learning from Adobe Stock's library, i.e. the assets uploaded by contributors, and b) compensated already for use of their assets in Firefly's training.
Why is Adobe only at the 'exploring' stage of an opt-out possibility and still busy 'developing' a compensation model? Shouldn't that be the first thing on the priority list?
Once again it looks like contributors and compensation of them are treated as an afterthought, a nuisance, an annoying expense.
-
To clarify, Kirsten is referencing the question from people who may not have received the actual email yesterday. As she stated, you must be opted in to receive marketing emails via adobe.com to receive this sort of information in real time.
You can be assured that any major announcements like the news about Firefly will be called out here in Microstock Group Forum too. :)
Thank you,
Mat Hayward
Getty images had their images pilfered to train an AI and they are in the process of mounting a significant legal action against the company that did so.
Shutterstock compensate it's contributers for images that have been used in the data sets to train their A.I. you are also not able to opt out.
Adobe is compensating their contributers how. I think that ignoring this question which has already been asked doesn't make for a good look. So what's the deal. This isn't light and fluffy ... awwww cummon guys ... this is business. You don't get free use. It is a requirement to be transparent regarding payment for use. what Matt.
There is an opt-out in the Shutterstock Account Settings>Licensing Options, although it does feel as if this is closing a door after the horse has bolted.
-
To clarify, Kirsten is referencing the question from people who may not have received the actual email yesterday. As she stated, you must be opted in to receive marketing emails via adobe.com to receive this sort of information in real time.
You can be assured that any major announcements like the news about Firefly will be called out here in Microstock Group Forum too. :)
Thank you,
Mat Hayward
Getty images had their images pilfered to train an AI and they are in the process of mounting a significant legal action against the company that did so.
Shutterstock compensate it's contributers for images that have been used in the data sets to train their A.I. you are also not able to opt out.
Adobe is compensating their contributers how. I think that ignoring this question which has already been asked doesn't make for a good look. So what's the deal. This isn't light and fluffy ... awwww cummon guys ... this is business. You don't get free use. It is a requirement to be transparent regarding payment for use. what Matt.
There is an opt-out in the Shutterstock Account Settings>Licensing Options, although it does feel as if this is closing a door after the horse has bolted.
Exactly. I've opted out, but site slike DALL-E or modjourney have long since used my images to train their AIs and even Shutterstock only added the opt out option after they had launched their AI generation option. It feels more like a ruse after all the damage has already been done.
-
@MatHayward
Is there any chance that there will be Ai search term in filters or category in search panel on site to help users / buyers decide what kind of artwork or "airtwork" they wish ?
Or at least something like a subcategory for people - people with 6 fingers, people with 7 fingers ...you know....something like that?
already there just include -ai in your search, since AS requires that keyword
And just a suspicion that agencies want AI images marked, so they won't be used to train AI.
-
Firefly has been fun to play with, and it's amazing that conceptually it gets the prompt to image pretty correct, but the actual results are atrocious in most cases. Just on basic stock things. Distorted balls, the extra or missing fingers we all love, garbled looking money, odd geometric snowflakes, etc.
-
I don't understand why contributors have not been ...,compensated already for use of their assets in Firefly's training.
Why is Adobe only at the 'exploring' stage of an opt-out possibility and still busy 'developing' a compensation model? Shouldn't that be the first thing on the priority list?
Once again it looks like contributors and compensation of them are treated as an afterthought, a nuisance, an annoying expense.
both AS & SS have the right to use those images under the TOS artists agreed to when contributing (everyone read those in detail, right?)
what they're doing NOW is allowing people to opt-out - they could have continued the existing policy that says contributors agree to those TOS
and still waiting for anyone to describe what they think would be fair compensation when their contribution swamp by hundreds of millions of images used for training
and also interesting that none of the loudest whingers (who claim their hard work creating art will be swamped by mediocre user generated AI art) will actually present their portfolios!
-
and also interesting that none of the loudest whingers (who claim their hard work creating art will be swamped by mediocre user generated AI art) will actually present their portfolios!
Maybe they don't, out of respect to yours
-
I don't understand why contributors have not been ...,compensated already for use of their assets in Firefly's training.
Why is Adobe only at the 'exploring' stage of an opt-out possibility and still busy 'developing' a compensation model? Shouldn't that be the first thing on the priority list?
Once again it looks like contributors and compensation of them are treated as an afterthought, a nuisance, an annoying expense.
both AS & SS have the right to use those images under the TOS artists agreed to when contributing (everyone read those in detail, right?)...
Both AS & SS claim they have the right to train their generative AI model with their stock library content. Their self interest clearly indicates why they'd say that. Claiming it to be so doesn't make it so - but there'd need to be legal action to get a court to agree that the agreement doesn't mean what the agencies claim.
I think the notion of using contributor content to assist in creating new Adobe Stock features (such as a keyword suggestion tool) is fine and reasonable - I did read the agreement and given my assumption that the agency and contributor both have aligned interests in increasing the licensing of agency content, had no concerns about it. Contractual deals that don't result in contributor royalties have been an issue for years - SS's deal with facebook and the more recent 5 year contracts to allow use of their library for AI training (https://finance.yahoo.com/m/3c6ebbdf-0128-3207-9c67-0d686a631666/shutterstock-has-found-an.html). The deals would not have happened without the content in their libraries but all the up front revenue goes straight to the agency.
I can't imagine contributors having the money to take a case to court, but I think it would be possible to argue that what Adobe has done does not fall within the scope of (from Adobe Stock's Contributor Agreement section 1) "...developing new features and services".
The tack I would take is that the context implied in the agreement is that the features and services would be part of the paid licensing of contributor works that is the scope of that agreement. What has been built with Firefly is (if it succeeds) potentially a replacement for the licensing of all of the works currently in the library.
I'd argue that no reasonable person (the legal concept of the reasonable man (https://www.law.cornell.edu/wex/reasonable_person)) reading the current agreement would think they were consenting to allow their works to be used to eliminate all future commercial value for those works. Imagine investor excitement at a revenue stream in the future that doesn't carry any royalty obligations - it would make SS's "margin optimization" efforts since June 2020 look like amateur hour.
As an aside, I don't see the questionable quality of AI generated content to be any barrier to market success - good enough wins out lots of times as long as the price is right and it's easy to access. When SS was promoting their AI-enhanced Canva-clone, the video tag line was "No time? No budget? No designer? No problem!". In other words, it may not matter that contributor work is "better" than what Firefly generates, hence the worries about competition.
I watched two interviews - one with Adobe's CEO Shantanu Narayen and the other with their President of Digital Media, David Wadhwani. Other than the expected buzzy optimism about their new offerings, there were a couple of interesting (from the Adobe Stock contributor perspective) comments. Adobe plans to offer an API to allow other companies to use their generative AI model - nothing was said about whether and how Adobe Stock contributors whose work was the training basis for the model would be compensated in that cases.
The CEO mentioned t hat it would be possible to have stock contributors supply their styles to Firefly customers (no timeline or compensation model mentioned). He also mentioned offering Adobe's model to large customers who wanted to add their own repository of data to have a customized generative AI - no notion of whether there would be royalties to Adobe Stock contributors there either. I didn't care for the way the CEO referred to "our" model trained by "our" content in talking about this future product offering. Content on Adobe Stock is not Adobe's content. The CEO said they refer to the content supply chain and how important it is to get that worked out.
The interviewer asked David Wadhwani what the incentive was for Adobe Stock contributors to let Adobe train the AI on our images. The reply was a set of reassuring words about how this would be commercially good for Adobe Stock contributors. He didn't mention the less appealing reality that they didn't ask, so contributors had no say in this - not even to delete their portfolios if they really didn't like it - because the first we heard about this was after the training was complete.
There was a follow up question about compensation, citing Spotify as an example of how artists got tiny amounts per stream, and there was another fuzzy answer about this being worked out in concert and in conversation with contributors. I don't know anything about contributors being asked about how they'd like to be paid, but Spotify's model wouldn't be it (where they pay the big name artists a larger share of the revenue to keep them happy and the smaller artists get very little).
There was talk about growing the number of Creative Cloud subscribers - bringing in new users (I hadn't heard the term solopreneurs before) increasing Adobe's Total Addressable Market (that's aimed at keeping investors happy). No date given when asked how long the beta would be. Beta is just images and text effects. They will be doing video, layout, design, 3D, brushes & styles.
Here are the videos
Firefly AI: I speak with Adobe CEO Shantanu Narayen from Adobe Summit
https://www.youtube.com/watch?v=04X6h2K2WDw (https://www.youtube.com/watch?v=04X6h2K2WDw)
Generative AI Comes to Adobe Creative Cloud: Adobe's David Wadhwani on Firefly
https://www.youtube.com/watch?v=v4dgIUdiZfQ (https://www.youtube.com/watch?v=v4dgIUdiZfQ)
-
I watched two interviews - one with Adobe's CEO Shantanu Narayen and the other with their President of Digital Media, David Wadhwani.
Thank you for the very detailed summery, you always do such a god job of these!
I laughed at the part where Wadhwan claimes there was anything being worked out in conversation with contributors, because we all very well know that this is not happening.
Notice how even here, whenever Mat announces something new, he will always just reply to questions and comments stricly related to the technical aspect of a feature and ignore every single question and comment about user actual concerns or morality aspects of what Adobe was doing.
I can't recall that Adobe has ever seeked out conversation with contributors about anything, before we were getting presented the end result - Free galleries, currency changes, our images being used to train AI that will make our images worthless - there has never been any kind of "conversation" prior to annoucing the final decision and the "conversation" that took place after that were alwas one-sided with Adobe pretty much ignoring our concerns.
-
I don't understand why contributors have not been ...,compensated already for use of their assets in Firefly's training.
Why is Adobe only at the 'exploring' stage of an opt-out possibility and still busy 'developing' a compensation model? Shouldn't that be the first thing on the priority list?
Once again it looks like contributors and compensation of them are treated as an afterthought, a nuisance, an annoying expense.
both AS & SS have the right to use those images under the TOS artists agreed to when contributing (everyone read those in detail, right?)
what they're doing NOW is allowing people to opt-out - they could have continued the existing policy that says contributors agree to those TOS
and still waiting for anyone to describe what they think would be fair compensation when their contribution swamp by hundreds of millions of images used for training
and also interesting that none of the loudest whingers (who claim their hard work creating art will be swamped by mediocre user generated AI art) will actually present their portfolios!
Exactly what Jo Ann says above: I agreed to let them use my images for things like a keyword suggestion tool, or for promotion of their library and assets, anything to increase revenue for the good of contributors.
Firefly however is created with the intention to replace our work and efforts in the long run, which I never signed up for (back in 2010 when I signed up at Fotolia high quality AI image creation wasn't even a feasible thing yet).
I'd argue this AI thing falls outside of the scope of that agreement and needs proper attention and royalty compensation.
-
hought, a ...
both AS & SS have the right to use those images under the TOS artists agreed to when contributing (everyone read those in detail, right?)...
Both AS & SS claim they have the right to train their generative AI model with their stock library content. Their self interest clearly indicates why they'd say that. Claiming it to be so doesn't make it so - but there'd need to be legal action to get a court to agree that the agreement doesn't mean what the agencies claim.
I think the notion of using contributor content to assist in creating new Adobe Stock features (such as a keyword suggestion tool) is fine and reasonable - I did read the agreement and given my assumption that the agency and contributor both have aligned interests in increasing the licensing of agency content, had no concerns about it. Contractual deals that don't result in contributor royalties have been an issue for years - SS's deal with facebook and the more recent 5 year contracts to allow use of their library for AI training (https://finance.yahoo.com/m/3c6ebbdf-0128-3207-9c67-0d686a631666/shutterstock-has-found-an.html). The deals would not have happened without the content in their libraries but all the up front revenue goes straight to the agency.
I can't imagine contributors having the money to take a case to court, but I think it would be possible to argue that what Adobe has done does not fall within the scope of (from Adobe Stock's Contributor Agreement section 1) "...developing new features and services".
The tack I would take is that the context implied in the agreement is that the features and services would be part of the paid licensing of contributor works that is the scope of that agreement. What has been built with Firefly is (if it succeeds) potentially a replacement for the licensing of all of the works currently in the library. ...
thanks yet again for another detailed examination of the issues (and of course i'd include you among those who did read the TOS
while i do agree with your analysis of the fuzziness of their claims and the possibility of a successful lawsuit, i focus on the actual expected returns to individual artists. what would actual payments to authors be? (and given how ML works any payment for use would only be microscopic as each sale would also be spread over the entire training set, and the legal grounds for that are even murkier)
my basic stance is since the chances of a successful legal action are remote (mostly due to the costs), i don't spend a lot of time worrying about things i can't change & concentrate on those over which i can change (bloggig, SEO, social media,etc)
-
How long does it take to get the approval to participate in the beta after submitting the request?
-
...
The tack I would take is that the context implied in the agreement is that the features and services would be part of the paid licensing of contributor works that is the scope of that agreement. What has been built with Firefly is (if it succeeds) potentially a replacement for the licensing of all of the works currently in the library. ...
100% agree. It's compounded when you read comments, under their YouTube video for example, where the major selling point for buyers is that this is a morally acceptable AI tool.
People are saying they were conflicted about using AI as it takes artists’ work without permission but they can at last use this because Adobe is training it on sets where this isn’t an issue. They belong to Adobe.
IMHO what Adobe has done by helping themselves to our work without compensation and far outside the scope of our contract and claiming it is “their” library as a major selling point is morally reprehensible. It’s the worst behaviour out of the three big libraries (Getty and SS being the other two).
-
Firefly has been fun to play with, and it's amazing that conceptually it gets the prompt to image pretty correct, but the actual results are atrocious in most cases. Just on basic stock things. Distorted balls, the extra or missing fingers we all love, garbled looking money, odd geometric snowflakes, etc.
Interesting. We tend to talk about AI as if its a monolith. In my experience the different engines have vastly different quality. Midjourney, for example, is light-years ahead of Dall E at the moment.
-
Notice how even here, whenever Mat announces something new, he will always just reply to questions and comments stricly related to the technical aspect of a feature and ignore every single question and comment about user actual concerns or morality aspects of what Adobe was doing.
I can't recall that Adobe has ever seeked out conversation with contributors about anything, before we were getting presented the end result - Free galleries, currency changes, our images being used to train AI that will make our images worthless - there has never been any kind of "conversation" prior to annoucing the final decision and the "conversation" that took place after that were alwas one-sided with Adobe pretty much ignoring our concerns.
Very true. And if you accuse Mat of corporate speak then he is offended :)
I do think they listen in via Mat. I actually think that Adobe didn't want to prolong the bonus program but because here, and probably at other places as well, we made a fuzz about it and so they caved in, in the end. Not to lose face and they sure will have weighted the costs versus the potential reputation damage.
But it annoys me too that when you ask something else then a technical question you never hear an answer. Concerns are maybe listened to but certainly not replied to.
And since they are a company (maximum profit seeking) it annoys me that they pretend to be an NGO sometimes only to support artists at all times while in reality they are not. But I guess lot of people fall for that kind of talk. At least here, I see many of them. But you can also see them when you join some of their live Behance meetings. Some actually seem groupies and completely believe in Adobes fantasy tale :)
And remember there is no them without us but not exactly as how they mean it. It's more that if we don't buy their products or deliver content to them then there is no them. That part is definitely true.
-
Notice how even here, whenever Mat announces something new, he will always just reply to questions and comments stricly related to the technical aspect of a feature and ignore every single question and comment about user actual concerns or morality aspects of what Adobe was doing.
I can't recall that Adobe has ever seeked out conversation with contributors about anything, before we were getting presented the end result - Free galleries, currency changes, our images being used to train AI that will make our images worthless - there has never been any kind of "conversation" prior to annoucing the final decision and the "conversation" that took place after that were alwas one-sided with Adobe pretty much ignoring our concerns.
Very true. And if you accuse Mat of corporate speak then he is offended :)
I do think they listen in via Mat. I actually think that Adobe didn't want to prolong the bonus program but because here, and probably at other places as well, we made a fuzz about it and so they caved in, in the end. Not to lose face and they sure will have weighted the costs versus the potential reputation damage.
But it annoys me too that when you ask something else then a technical question you never hear an answer. Concerns are maybe listened to but certainly not replied to.
And since they are a company (maximum profit seeking) it annoys me that they pretend to be an NGO sometimes only to support artists at all times while in reality they are not. But I guess lot of people fall for that kind of talk. At least here, I see many of them. But you can also see them when you join some of their live Behance meetings. Some actually seem groupies and completely believe in Adobes fantasy tale :)
And remember there is no them without us but not exactly as how they mean it. It's more that if we don't buy their products or deliver content to them then there is no them. That part is definitely true.
Yup, I'm so glad somebody said that. I see it too and often feel like traitor if I dare to say something against Adobe. Because looks that many people think they really care for us.
-
Still waiting for my invite :(
-
Notice how even here, whenever Mat announces something new, he will always just reply to questions and comments stricly related to the technical aspect of a feature and ignore every single question and comment about user actual concerns or morality aspects of what Adobe was doing.
I can't recall that Adobe has ever seeked out conversation with contributors about anything, before we were getting presented the end result - Free galleries, currency changes, our images being used to train AI that will make our images worthless - there has never been any kind of "conversation" prior to annoucing the final decision and the "conversation" that took place after that were alwas one-sided with Adobe pretty much ignoring our concerns.
Very true. And if you accuse Mat of corporate speak then he is offended :)
I do think they listen in via Mat. I actually think that Adobe didn't want to prolong the bonus program but because here, and probably at other places as well, we made a fuzz about it and so they caved in, in the end. Not to lose face and they sure will have weighted the costs versus the potential reputation damage.
But it annoys me too that when you ask something else then a technical question you never hear an answer. Concerns are maybe listened to but certainly not replied to.
And since they are a company (maximum profit seeking) it annoys me that they pretend to be an NGO sometimes only to support artists at all times while in reality they are not. But I guess lot of people fall for that kind of talk. At least here, I see many of them. But you can also see them when you join some of their live Behance meetings. Some actually seem groupies and completely believe in Adobes fantasy tale :)
And remember there is no them without us but not exactly as how they mean it. It's more that if we don't buy their products or deliver content to them then there is no them. That part is definitely true.
Well, I'm not a fan of any corporation pretending to be a "community of artists", but I have nothing against Mat personally. He is just an Adobe employee, plain and simple: a PR job and a messenger of good and bad news; if it's good news, he can speak freely with enthusiasm and happily answer questions, if it's "bad" news he will (have to) stick to the "press release" and whatever Adobe tells him to say. Which I understand, he cannot make any promises without permission from upper management or speak against his employer.
But it's better than complete radio silence.
-
"We are developing a compensation model for Stock contributors"
definition of compensation noun (MONEY): money that is paid to someone in exchange for something that has been lost or damaged or for some problem
Seems to be there is (for Stock contributors), a prejudice confessed here... ::)
-
Notice how even here, whenever Mat announces something new, he will always just reply to questions and comments stricly related to the technical aspect of a feature and ignore every single question and comment about user actual concerns or morality aspects of what Adobe was doing.
I can't recall that Adobe has ever seeked out conversation with contributors about anything, before we were getting presented the end result - Free galleries, currency changes, our images being used to train AI that will make our images worthless - there has never been any kind of "conversation" prior to annoucing the final decision and the "conversation" that took place after that were alwas one-sided with Adobe pretty much ignoring our concerns.
Very true. And if you accuse Mat of corporate speak then he is offended :)
I do think they listen in via Mat. I actually think that Adobe didn't want to prolong the bonus program but because here, and probably at other places as well, we made a fuzz about it and so they caved in, in the end. Not to lose face and they sure will have weighted the costs versus the potential reputation damage.
But it annoys me too that when you ask something else then a technical question you never hear an answer. Concerns are maybe listened to but certainly not replied to.
And since they are a company (maximum profit seeking) it annoys me that they pretend to be an NGO sometimes only to support artists at all times while in reality they are not. But I guess lot of people fall for that kind of talk. At least here, I see many of them. But you can also see them when you join some of their live Behance meetings. Some actually seem groupies and completely believe in Adobes fantasy tale :)
And remember there is no them without us but not exactly as how they mean it. It's more that if we don't buy their products or deliver content to them then there is no them. That part is definitely true.
Well, I'm not a fan of any corporation pretending to be a "community of artists", but I have nothing against Mat personally. He is just an Adobe employee, plain and simple: a PR job and a messenger of good and bad news; if it's good news, he can speak freely with enthusiasm and happily answer questions, if it's "bad" news he will (have to) stick to the "press release" and whatever Adobe tells him to say. Which I understand, he cannot make any promises without permission from upper management or speak against his employer.
But it's better than complete radio silence.
I never said there is something wrong with Mat personally. I actually think he is very likeable guy. It's only that he presents himself as a person that is here to speak on behalve of Adobe and you can discuss anything with him. But it ends up with only answering easy technical questions or singing hallelujah with us when there is good news to share. I do understand perfectly that he cannot answer the more difficult answers thrown at him but it does always stays quiet afterwards even when he would have had the chance to discuss it above. If he is the bringer of good news and/or bad news, again on behalve of Adobe, he should be able to respond (as an employee of Adobe) to all questions. Not only the easy ones or just the technical ones. Otherwise he is just being used as a PR machine and not as an actual liason of Adobe that can also help with our concerns about certain aspects.
It's about the presentation and the abuse of misconception of easy to fool persons here.
But I would drink a beer anytime with Mat. I can actually endorse all of his Behance meetings although you have to withstand the, over the top, American, hallelujah Adobe family b*s*t.
-
"We are developing a compensation model for Stock contributors"
definition of compensation noun (MONEY): money that is paid to someone in exchange for something that has been lost or damaged or for some problem
Seems to be there is (for Stock contributors), a prejudice confessed here... ::)
Don't be so deep?
NORTH AMERICAN ENGLISH:
the money received by an employee from an employer as a salary or wages.
or
Typically, compensation refers to monetary payment given to an individual in exchange for their services. In the workplace, compensation is what is earned by employees. It includes salary or wages in addition to commission and any incentives or perks that come with the given employee’s position.
Now what about the incentives and perks part? For example, maybe a free subscription for a year to Creative Cloud software is the future compensation? ;D
-
Derivative Work Under Copyright Law
§ 106(2)). It is considered copyright infringement to make or sell derivative works without permission from the original owner, which is where licenses typically come into play.2 Jun 2017
Derivative work is clearly what an AI is producing. in which case compensation isn't a nice thing to do it is a legal requirement. Adobe have not paid for any such license to use the work. They don't own the copywrite we do.
Derivative work has its own copywrite. However who wons that. The A.I. or the A.I.s owner. And is it actually protected under copywrite because it is a Derivative work from copywritten material which wasn't paid for ...
-
Derivative work is clearly what an AI is producing. in which case compensation isn't a nice thing to do it is a legal requirement. Adobe have not paid for any such license to use the work. They don't own the copywrite we do.
..
not so clear - it's been shown many times that the actual generative AI does not use any images directly but creates an entirely new image from the training set.
whether anyone has the right to use images scraped from the web is a separate issue, more theoretical, since any payment to authors for the training would be minuscule a tiny fraction of 100s of m
at the same time writers & journalists arent complaining about scraping for trillions image examined
dont know why some artists continue to propagate misleading info especially since there's really no upside.
webscraping creates the dataset but the results of those GPT bots do not violate anyone's copyright, again because the AI generates completely new text w/o using the training data
-
Derivative work is clearly what an AI is producing. in which case compensation isn't a nice thing to do it is a legal requirement. Adobe have not paid for any such license to use the work. They don't own the copywrite we do.
It is clearly not what it is producing. And this is obvious to anyone who has tried an AI tool.
AI learns the same way as a human scrolling through google or pinterest and "getting inspired" with ideas of others, not paying a dime for "seeing, "memorising" and "collecting visual data" from copyrighted works.
Why haven't you paid for every image you have looked at today? It is clearly stored in your memory now and can (will! even if for 0.0001%) influence your creations. So, please pay for every piece of art, photo, design, furniture piece that surrounds you. What a mad world that would be? :)
AI does the same. It doesn't copy-paste anything at all, so nothing "derivative" there.
That is why copyright nonsense by stock sites doesn't make any sense, other than them trying to save their own existence. By being able to sue everything under the sun as Getty does.
Because following this logic all works created by people are also "derivative" then because artist creates them after he got inspired by a mix of many kinds of "copyrighted" things he has seen before! Nothing ever is born in the vacuum.
-
Because following this logic all works created by people are also "derivative" then because artist creates them after he got inspired by a mix of many kinds of "copyrighted" things he has seen before! Nothing ever is born in the vacuum.
Wrong.
It's a long-standing law principle that copyright does not protect ideas, concepts, systems, or methods of doing something, while it differentiates all this from derivative work.
Obviously you are not a lawyer and you didn't do your homework.
-
Because following this logic all works created by people are also "derivative" then because artist creates them after he got inspired by a mix of many kinds of "copyrighted" things he has seen before! Nothing ever is born in the vacuum.
Wrong.
It's a long-standing law principle that copyright does not protect ideas, concepts, systems, or methods of doing something, while it differentiates all this from derivative work.
Obviously you are not a lawyer and you didn't do your homework.
This is hilarious how you seem to be stalking my every message to say it is wrong without reading it :D
Where did I say that copyright protects ideas?
Even that quoted by you sentence starts with words: following this logic (which is flawed as I described above)... etc. meaning that if it would be true it would be absurd, which is exactly what you have repeated after me but said that it is wrong :D
Sorry, I cannot even follow your reply, because you reply on something I wasn't even talking about :D
If I need to repeat especially for you in a simple sentence: AI doesn't create derivative work, same as artist doesn't create derivative work, because memorising, learning and being inspired by something to create something new is not same as copy-pasting.
What are you talking about I don't know :)
-
Because following this logic all works created by people are also "derivative" then because artist creates them after he got inspired by a mix of many kinds of "copyrighted" things he has seen before! Nothing ever is born in the vacuum.
Wrong.
It's a long-standing law principle that copyright does not protect ideas, concepts, systems, or methods of doing something, while it differentiates all this from derivative work.
Obviously you are not a lawyer and you didn't do your homework.
This is hilarious how you seem to be stalking my every message to say it wrong without reading it :D
Where did I say that copyright protects ideas?
Even that quoted by you sentence starts with words: following this logic (which is flawed as I described above)... etc.
Sorry, I cannot even follow your reply, because you reply on something I wasn't even talking about :D
If I need to repeat especially for you in a simple sentence: AI doesn't create derivative work, same as artist doesn't create derivative work, because memorising, learning and being inspired by something is not same as copy-pasting.
What are you talking about I don't know :)
Nah, check your statements. You were ranting about Getty, calling their copyright lawsuits nonsense, without understanding the difference between ideas, concepts, and derivative work, assuming that these are equivalent.
Here is a clear example of an AI generated image that is not just an idea, but clearly derivative work:
(https://www.microstockgroup.com/shutterstock-com/shutterstock-'contributor-fund'/?action=dlattach;attach=19298;image)
... and the paper behind the study the example is coming from: https://arxiv.org/pdf/2301.13188.pdf (https://arxiv.org/pdf/2301.13188.pdf)
PS. rest assured that you are not the belly button of the world. ;) My reply was not about you, but about the error, you made in your post.
-
Repeating my sentence: "That is why copyright nonsense by stock sites doesn't make any sense, other than them trying to save their own existence. By being able to sue everything under the sun as Getty does."
COPYRIGHT NONSENSE REGARDING AI CONTENT, WE ARE IN THE THREAD ABOUT AI.
Sorry, it is very hard to convey ideas to someone who takes things out of contex and drives them into his own conversations with himself.
-
Repeating my sentence: "That is why copyright nonsense by stock sites doesn't make any sense, other than them trying to save their own existence. By being able to sue everything under the sun as Getty does."
COPYRIGHT NONSENSE REGARDING AI CONTENT, WE ARE IN THE THREAD ABOUT AI.
Sorry, it is very hard to convey ideas to someone who takes things out of contex and drives them into his own conversations with himself.
Yep, "everything under the sun".
Obviously, you don't understand why Getty is suing (hence protecting their business and OUR work) assuming that their lawsuits are frivolous!
Getty has a problem with AI companies using their database and OUR work without permission.
Besides, the example and the white paper I showed you is another proof that you are wrong: when the training set is limited, AI may very well end up plagiarizing copyrighted work. So it doesn't generate only ideas and concepts, but also derivative work.
-
Yep, "everything under the sun".
Obviously, you don't understand why Getty is suing (hence protecting their business and OUR work) assuming that their lawsuits are frivolous!
Getty has a problem with AI companies using their database and OUR work without permission.
Besides, the example and the white paper I showed you is another proof that you are wrong: when the training set is limited, AI may very well end up plagiarizing copyrighted work. So it doesn't generate only ideas and concepts, but also derivative work.
You are correct. The people disagreeing tend to be the ones who stand to make money out of the new technology.
If the AI has a dataset of one the output will be (near) identical. Two and it will strongly resemble the two images, and so on. Once you get to millions it is a lot harder to spot but the principle is still very much the same. Infringing 1000,000 people’s copyright isn’t better than infringing one’s. It is a lot worse because it destroys an industry and many jobs. If I steal one person’s images we may not even end up competing for the same customers. If I come up with an AI engine to steal everyone’s I have destroyed artists’ ability to make a living.
Anthropomorphising the app is silly. It isn’t “learning” in the same way as a human. It’s taking and processing images via programming. Saying they don’t use them “directly” is a bizarre concept when we are talking about images being loaded into any type of algorithm or app resulting in “new” images coming out the other end.
The law is very clear. The Berne convention says the following must to be recognised as exclusive rights of authorisation. “the right to make adaptations and arrangements of the work”. Reading the exceptions makes the sprit of the law even clearer. The theme is exceptions are only allowed if they don’t conflict with a normal exploitation of the work and don't prejudice the legitimate interests of the author. AI very clearly does both.
Just to add, don’t read this as me being optimistic as to the outcome of any legal claims. The law is there to serve the powerful and all the big tech companies are behind the tech. It’s unstoppable (not because it is currently legal or “right” though). I expect the law to be changed to allow it for these reasons.
ETA: missed this bit. Author also has exclusive "right to use the work as a basis for an audiovisual work". Not sure how you can argue that the images ingested haven't been used as the "basis" for the AI's output, regardless of how much/ little, directly/ indirectly they are processed. No image inputs, no image outputs.
-
...
Getty has a problem with AI companies using their database and OUR work without permission.
...
To add one more thing as this is an Adobe thread. Adobe helping themselves to work we uploaded for completely different reasons is even worse IMHO (especially as they marketing it as a more "ethical" AI as it only uses "their" images).
-
Because following this logic all works created by people are also "derivative" then because artist creates them after he got inspired by a mix of many kinds of "copyrighted" things he has seen before! Nothing ever is born in the vacuum.
Wrong.
It's a long-standing law principle that copyright does not protect ideas, concepts, systems, or methods of doing something, while it differentiates all this from derivative work.
Obviously you are not a lawyer and you didn't do your homework.
but that's the whole point! AI work is not derivative in an rational sense & no one has yet given an example that shows such
-
..
Besides, the example and the white paper I showed you is another proof that you are wrong: when the training set is limited, AI may very well end up plagiarizing copyrighted work. So it doesn't generate only ideas and concepts, but also derivative work.
another strawwoman argument - the discussion was about massive AI training sets - of course giving it a limited dataset can produce what may be derivative work - but even submitting single images often produce unrecognizable results
-
...
Anthropomorphising the app is silly. It isn’t “learning” in the same way as a human. It’s taking and processing images via programming. Saying they don’t use them “directly” is a bizarre concept when we are talking about images being loaded into any type of algorithm or app resulting in “new” images coming out the other end.
...
MOST successful AI doesn't learn like a human but produces vastly superior results - champion GO & chess Ai dont even know the rules of the game! it's called emergent behavir
-
of course giving it a limited dataset can produce what may be derivative work
Exactly!
-
of course giving it a limited dataset can produce what may be derivative work
Exactly!
Let's take the worst case non-existent scenario and judge it all based on it.
Let's ban all the kitchen knives because very occasionally someone is got killed with them.
The talk was about legitimate companies that allow commercial use because they are sure such cases are too rare (and their goal is to bring them to zero eventually, it's been few months!), not some garage AI.
And even in that overblown unimaginable case with 1 input of data, it may not be creating a derivative, if you have to draw a tree there is no way you can draw it any other way than in a shape of a tree. And if you've seen only one tree ever then well what would you expect. But it doesn't copy-paste nor collage anything - this is not how it works at all.
-
Because following this logic all works created by people are also "derivative" then because artist creates them after he got inspired by a mix of many kinds of "copyrighted" things he has seen before! Nothing ever is born in the vacuum.
Wrong.
It's a long-standing law principle that copyright does not protect ideas, concepts, systems, or methods of doing something, while it differentiates all this from derivative work.
Obviously you are not a lawyer and you didn't do your homework.
This is hilarious how you seem to be stalking my every message to say it is wrong without reading it :D
Where did I say that copyright protects ideas?
Even that quoted by you sentence starts with words: following this logic (which is flawed as I described above)... etc. meaning that if it would be true it would be absurd, which is exactly what you have repeated after me but said that it is wrong :D
Sorry, I cannot even follow your reply, because you reply on something I wasn't even talking about :D
If I need to repeat especially for you in a simple sentence: AI doesn't create derivative work, same as artist doesn't create derivative work, because memorising, learning and being inspired by something to create something new is not same as copy-pasting.
What are you talking about I don't know :)
you are clearly delusional. Derivative Work is literally an accepted technique and has its own copywrite law. I won't embarrass you any more other than to state that if it was nonsense as you suggest SS wouldn't pay people for the use of their images in the learning sets because it is 'only storing' Our images as reference and yet SS is paying contributers. Getty wouldn't have a legal case going and yet they do. And Adobe wouldn't be exploring a compensation model and yet they are. Evidence and random ranter on a forum. I know what I believe ... but please ... continueninnyour delusion just do so quietly please.
-
Derivative work is clearly what an AI is producing. in which case compensation isn't a nice thing to do it is a legal requirement. Adobe have not paid for any such license to use the work. They don't own the copywrite we do.
..
not so clear - it's been shown many times that the actual generative AI does not use any images directly but creates an entirely new image from the training set.
whether anyone has the right to use images scraped from the web is a separate issue, more theoretical, since any payment to authors for the training would be minuscule a tiny fraction of 100s of m
at the same time writers & journalists arent complaining about scraping for trillions image examined
dont know why some artists continue to propagate misleading info especially since there's really no upside.
webscraping creates the dataset but the results of those GPT bots do not violate anyone's copyright, again because the AI generates completely new text w/o using the training data
Well you are conflating two issues there.
firstly the data sets it uses are an unknown quantity so you and others are guessing. Assuming millions is a nonsense because it just doesnt need to. Not when it could use 5.It doesnt work like that. As others have stated and shown examples of that given a limited data set the image will be similar or because it isnt programmed to be innovative it will or rather cab directly duplicate it. I've seen many examples of A.I. generated images and they have created a clearly derived work. And another issue is that asking it to create a photo of a dog in a costume should generate a fairly generic photo. But users don't do that. They'll ask for a specific breed, in a specific costume in a specific environment which limits the data sets it can reference. Similarly ask it to create a dragonfly on some grasses and it will reference only a few images not all.
Your second example of text regarding ChatGBT4 producing unique text is not quite correct. This is new law in the making. plagiarism has a long standing legal framework but even though ChatGBt4 can create text with perimeters (write an essay about World War 2 as if I had written it) producing a worryingly me like essay. But again it has racesd ahead and is being abusesd but not so well publicised is that apps have been created to detect A.I. created text already. Because of the potential for plagiarism. Schools are trialing A.I. detection software as we speak using A.I. to find content produced by a.I.
It is a very new frontier across the spectrum of usage but shortly the law will catch up as it always does. And companies that played fast and loose will find themselves in a very uncomfortable position of compensation and fines. Which is why SS are getting ahead of the curve by at least making a show of being wholesome. Random payments for images used that they do not disclose and payments are extremely variable, which doesnt help with transparency but, they are acknowledging that compensation must be paid.
Adobe will have been aware of this and I suspect have already stated "we are working on a model of compensation" because the A.I. output isn't so far from its source material as they were led to believe it would be so are playing catch up with compensation. Just a theory of course.
-
of course giving it a limited dataset can produce what may be derivative work
Exactly!
Let's take the worst case non-existent scenario and judge it all based on it.
Let's ban all the kitchen knives because very occasionally someone is got killed with them.
The talk was about legitimate companies that allow commercial use because they are sure such cases are too rare (and their goal is to bring them to zero eventually, it's been few months!), not some garage AI.
And even in that overblown unimaginable case with 1 input of data, it may not be creating a derivative, if you have to draw a tree there is no way you can draw it any other way than in a shape of a tree. And if you've seen only one tree ever then well what would you expect. But it doesn't copy-paste nor collage anything - this is not how it works at all.
No, it's not a non-existing scenario.
This scenario is very much valid for those contributors who create really unique content. If their unique content is allowed in the AI training set, then the AI output will simply be plagiarism.
No problem with sliced tomatoes, hamburgers, or business people shaking hands, if that's your domain.
On the other hand, if you don't just do business people handshaking hands, sliced tomatoes or hamburgers, but truly unique content, then it's in your best interest to opt out of any AI training scheme (even if they pay you 1 peanut), to avoid shooting yourself in the foot.
And it's in the interest of all microstock companies to prevent this unique content to be used for AI training, or else they may face plagiarism lawsuits.
-
Researchers only extracted 94 direct matches and 109 perceptual near-matches out of 350,000 high-probability-of-memorization images they tested (a set of known duplicates in the 160 million-image dataset used to train Stable Diffusion), resulting in a roughly 0.03 percent memorization rate in this particular scenario.
Also, the researchers note that the "memorization" they've discovered is approximate since the AI model cannot produce identical byte-for-byte copies of the training images. By definition, Stable Diffusion cannot memorize large amounts of data because the size of the 160 million-image training dataset is many orders of magnitude larger than the 2GB Stable Diffusion AI model. That means any memorization that exists in the model is small, rare, and very difficult to accidentally extract.
None of this is written by me :D
@Lowls Go and write your insulting replies under this article: https://arstechnica.com/information-technology/2023/02/researchers-extract-training-images-from-stable-diffusion-but-its-difficult/amp/
Or maybe straight to a group of AI researchers from Google, DeepMind, UC Berkeley, Princeton, and ETH Zurich who released this paper, they are very interested in your opinion.
(Since I ignore you from now on, so won't see your little insulting replies anyway. :) )
Thanks to the forum for this feature.
-
When training an image synthesis model, researchers feed millions of existing images into the model from a dataset, typically obtained from the public web. The model then compresses knowledge of each image into a series of statistical weights, which form the neural network. This compressed knowledge is stored in a lower-dimensional representation called "latent space." Sampling from this latent space allows the model to generate new images with similar properties to those in the training data set.
This is how AI works. (not written by me either :D)
Surprisingly nothing about derivatives and copy-pasting... Probably because they don't read this thread well enough to follow all the expert advice properly :)
-
Sorry this has got heated/ personal. There's no need for it.
Back on topic.
Or maybe straight to a group of AI researchers...
Possibly not the best people to trust on the topic as they have literally the most invested in it being legit and above board. Even then the paper produces examples of the AI copying images almost exactly. In fact a lot more exactly than if I just saved the exact same image in a more lossy file format.
Surprisingly nothing about derivatives and copy-pasting...
I mean is there nothing about it? Strip out the buzzwords and leading language it could mean almost anything. Here's a possible translation:
When training an image synthesis model, researchers feed millions of existing images into the model from a dataset, typically obtained from the public web.
When programming an AI app employees copy millions of our illustrations/ photos into the app, typically “taken” from our online portfolios where we posted them for completely unrelated reasons.
The model then compresses knowledge of each image into a series of statistical weights, which form the neural network.
The app then saves the images in a more compressed format and indexes them for easy retrieval when recompiling derivates.
This compressed knowledge is stored in a lower-dimensional representation called "latent space."
These compressed copies of the images are stored in a database.
Sampling from this latent space allows the model to generate new images with similar properties to those in the training data set.
Copying elements from the images in this format allows the app to create the derivative images using the properties of the original images compressed and indexed in its database.
OR:
When programming an AI app employees copy millions of our illustrations/ photos into the app, typically “taken” from our online portfolios where we posted them for completely unrelated reasons. The app then saves the images in a more compressed format and indexes them for easy retrieval when recompiling derivates. These compressed copies of the images are stored in a database. Copying elements from the images in this format allows the app to create the derivative images using the properties of the original images compressed and indexed in its database.
I am sure you can imagine writing a description of saving a jpeg as a tif and then zipping the file and making it sound like magic forming a new image if you used enough tech bro buzz words in a similar way.
-
Just look at the production of some AI advocates.
Look at the images they made before AI, and the images produced with AI that they proudly show off as their integral creation. Easy to check, since they expose a link to their portfolio or their website. Or go on Adobe stock and click on those latest AI stuff, and see the portfolio and what kind of production did the person before.
This simple fact is enough to understand what is going on. We can understand the temptation and the intoxicating feeling of having suddenly become great artists.
And some have (or will have) no qualms (and no state of mind) about flooding microstock sites with "their" art (of clicking a "generate" button).
...and they even complain that "Midjourney can legally sublicense any assets produced using it"... it's a competition of the unscrupulous!!!
What they don't realize is that they are only useful for this transition period, to select and check the images, they will be useless and discarded afterwards. The image generations will be autonomous, or directly driven by the final client. And this in a very short time, see you tomorrow!!! End of story.
-
What they don't realize is that they are only useful for this transition period, to select and check the images, they will be useless and discarded afterwards. The image generations will be autonomous, or directly driven by the final client. And this in a very short time, see you tomorrow!!! End of story.
Exactly! :) (though haven't seen anyone on this forum who doesn't realises it).
It is senseless to submit images to stock sites anymore (AI or not, maybe if only Editorial) as they themselves will cease to exist eventually, except Adobe as they have their own unique product. In my humble opinion (note for those who doesn't recognise it).
Some people here have predicted that it may happen within a year or even less.
-
We must anticipate the inabilities of AI, thanks to HI...
-
Notice how even here, whenever Mat announces something new, he will always just reply to questions and comments stricly related to the technical aspect of a feature and ignore every single question and comment about user actual concerns or morality aspects of what Adobe was doing.
I can't recall that Adobe has ever seeked out conversation with contributors about anything, before we were getting presented the end result - Free galleries, currency changes, our images being used to train AI that will make our images worthless - there has never been any kind of "conversation" prior to annoucing the final decision and the "conversation" that took place after that were alwas one-sided with Adobe pretty much ignoring our concerns.
Very true. And if you accuse Mat of corporate speak then he is offended :)
I do think they listen in via Mat. I actually think that Adobe didn't want to prolong the bonus program but because here, and probably at other places as well, we made a fuzz about it and so they caved in, in the end. Not to lose face and they sure will have weighted the costs versus the potential reputation damage.
But it annoys me too that when you ask something else then a technical question you never hear an answer. Concerns are maybe listened to but certainly not replied to.
And since they are a company (maximum profit seeking) it annoys me that they pretend to be an NGO sometimes only to support artists at all times while in reality they are not. But I guess lot of people fall for that kind of talk. At least here, I see many of them. But you can also see them when you join some of their live Behance meetings. Some actually seem groupies and completely believe in Adobes fantasy tale :)
And remember there is no them without us but not exactly as how they mean it. It's more that if we don't buy their products or deliver content to them then there is no them. That part is definitely true.
Yup, I'm so glad somebody said that. I see it too and often feel like traitor if I dare to say something against Adobe. Because looks that many people think they really care for us.
You see.
Adobe (via Mat) started this thread but all difficult questions were not answered. Even after some time, answers will not come.
But they are here for you, don't forget :)
-
Adobe just take your images, feed a machine with it and generates a lot money.
Did you contributors get even a penny for such a usage?
-
Adobe just take your images, feed a machine with it and generates a lot money.
Did you contributors get even a penny for such a usage?
Not sure I understand what you wrote but yes, we got paid, and yes it was pretty much pennies. I don't know how much we got per image, how many of mine, or what images were used. I may have missed that detail?
-
Adobe just take your images, feed a machine with it and generates a lot money.
Did you contributors get even a penny for such a usage?
No - there's a statement that they'll work out a compensation model and share the details when Firefly exits beta (it's still in beta now)
https://helpx.adobe.com/stock/contributor/help/firefly-faq-for-adobe-stock-contributors.html
-
No - there's a statement that they'll work out a compensation model and share the details when Firefly exits beta (it's still in beta now)
https://helpx.adobe.com/stock/contributor/help/firefly-faq-for-adobe-stock-contributors.html
Aight, and thanks. I was just curious about it.
At least, Adobe plans some sort of compensation, unlike the rest of the industry which just stealing images to train their machines.
I myself currently stopped uploading stuff to agencies until it's more clear that my rights are respected (and my images will be safe) or i get paid at least an extended licence per image.
-
What do you mean with "unlike the rest of the industry which just stealing images to train their machines" ?
Contributors at adobe have no choice.
Ah yes, one can delete portfolio, ok, really?
-if your content is used to train an AI model, it may not be possible to make the AI forget any learnings from your item.
-
I decided this should be its own topic
https://www.microstockgroup.com/fotolia-com/googleadobe-partnership-with-bard-firefly/ (https://www.microstockgroup.com/fotolia-com/googleadobe-partnership-with-bard-firefly/)
-
This is only available in the beta version of Photoshop (and I have no idea how one gets that):
https://techcrunch.com/2023/05/23/adobe-brings-fireflys-generative-ai-to-photoshop/ (https://techcrunch.com/2023/05/23/adobe-brings-fireflys-generative-ai-to-photoshop/)
"Photoshop is getting an infusion of generative AI today with the addition of a number of Firefly-based features that will allow users to extend images beyond their borders with Firefly-generated backgrounds, use generative AI to add objects to images and use a new generative fill feature to remove objects with far more precision than the previously available content-aware fill."
https://www.theverge.com/2023/5/23/23734027/adobe-photoshop-generative-fill-ai-image-generator-firefly (https://www.theverge.com/2023/5/23/23734027/adobe-photoshop-generative-fill-ai-image-generator-firefly)
According to Gizmodo, "Adobe’s Firefly Image Generator Is Going to Make Photoshop Much Easier to Use. Soon, even your grandparents could be Photoshop experts"
https://gizmodo.com/adobe-firefly-ai-image-generator-photoshop-generative-f-1850462988 (https://gizmodo.com/adobe-firefly-ai-image-generator-photoshop-generative-f-1850462988)
https://arstechnica.com/information-technology/2023/05/adobe-photoshops-new-generative-fill-ai-tool-lets-you-manipulate-photos-with-text/ (https://arstechnica.com/information-technology/2023/05/adobe-photoshops-new-generative-fill-ai-tool-lets-you-manipulate-photos-with-text/)
https://www.businessinsider.com/sc/adobe-photoshop-unlocks-a-new-era-of-generative-creativity-with-firefly (https://www.businessinsider.com/sc/adobe-photoshop-unlocks-a-new-era-of-generative-creativity-with-firefly)
https://9to5mac.com/2023/05/23/photoshop-first-adobe-app-with-generative-ai/ (https://9to5mac.com/2023/05/23/photoshop-first-adobe-app-with-generative-ai/)
https://www.digitalcameraworld.com/news/adobe-integrates-fireflys-generative-ai-with-photoshop (https://www.digitalcameraworld.com/news/adobe-integrates-fireflys-generative-ai-with-photoshop)
Still nothing more about the compensation model for the Adobe Stock contributors without whom none of this would be in beta...
-
This is only available in the beta version of Photoshop (and I have no idea how one gets that):
It's under "Beta Apps" in Creative Cloud Desktop App. Looks that it runs parallel with "normal" PS.
-
I understand Firefly Beta is not for commercial use. What about Photoshop Beta's new generative fill tool. Can we use it now to expand or alter our existing photos to create commercially uploadable (Adobestock) content? Or do we need to wait?
This tutorial shows the technique I'm referring to: https://www.youtube.com/watch?v=NvUZIm083P8 (https://www.youtube.com/watch?v=NvUZIm083P8)
Thanks
-
I understand Firefly Beta is not for commercial use. What about Photoshop Beta's new generative fill tool. Can we use it now to expand or alter our existing photos to create commercially uploadable (Adobestock) content? Or do we need to wait?
This tutorial shows the technique I'm referring to: https://www.youtube.com/watch?v=NvUZIm083P8 (https://www.youtube.com/watch?v=NvUZIm083P8)
Thanks
Thanks for the question. Firefly is still in beta which means it cannot be used commercially. Generative fill in Photoshop is a part of Firefly so you cannot use it on images for stock at this time.
thanks again,
Mat Hayward
-
More Firefly announcements yesterday - it's still in beta, but the emphasis on Adobe's rights to the content on which Firefly was trained, and the legal protection Adobe will offer customers, continue to be front and center of their marketing messages
https://www.reuters.com/technology/adobe-pushes-firefly-ai-into-big-business-with-financial-cover-2023-06-08/ (https://www.reuters.com/technology/adobe-pushes-firefly-ai-into-big-business-with-financial-cover-2023-06-08/)
https://www.fastcompany.com/90906560/adobe-feels-so-confident-its-firefly-generative-ai-wont-breach-copyright-itll-cover-your-legal-bills (https://www.fastcompany.com/90906560/adobe-feels-so-confident-its-firefly-generative-ai-wont-breach-copyright-itll-cover-your-legal-bills)
"Adobe Firefly, the software giant’s AI-powered image generation and expansion tool, is being rolled out to businesses today. At its flagship Adobe Summit event, the company is unveiling an expansion of Firefly for enterprise users that will include “full indemnification for the content created through these features,”"
"The Firefly model is trained on stock images for which Adobe already holds the rights..."
" “Adobe claim that Firefly has been trained on entirely legal inputs, mostly from their own extensive image libraries,” says Andres Guadamuz, an intellectual property law researcher at the University of Sussex. “This is an indication that they have conducted a thorough investigation of their training sources and are happy that they will not get sued.” Guadamuz adds that Adobe’s promise shouldn’t be taken lightly: “They must have some very strong assurances from their legal team that they’re in the clear,” he says. “I can’t imagine that they would do this if there was some doubt that they would get sued out of existence.”"
https://www.theverge.com/2023/6/8/23753564/adobe-firefly-enterprise-generative-ai-express-commercial (https://www.theverge.com/2023/6/8/23753564/adobe-firefly-enterprise-generative-ai-express-commercial)
"Adobe created Firefly to be safe for commercial use by training it on Adobe Stock images, openly licensed content, and content without copyright restrictions within the public domain."
https://techcrunch.com/2023/06/08/adobe-brings-firefly-to-the-enterprise/ (https://techcrunch.com/2023/06/08/adobe-brings-firefly-to-the-enterprise/)
"The major advantage that Adobe has been banking on since the launch of Firefly is that it produces commercially safe images. It’s training the model on images from its stock imagery marketplace (in addition to openly licensed images and public domain content), which means it has the rights to all of these images and doesn’t have to scrape the web to gather them, with all of the copyright issues that entails. "
https://www.forbes.com/sites/rashishrivastava/2023/06/08/adobe-brings-its-generative-ai-tool-firefly-to-businesses/?sh=48313750582b (https://www.forbes.com/sites/rashishrivastava/2023/06/08/adobe-brings-its-generative-ai-tool-firefly-to-businesses/?sh=48313750582b)
"Firefly is trained on more than 100 million images including Adobe’s stock images, licensed images and public images whose copyrights have expired. The company relies on its treasure trove of high-quality stock images sourced from contributors who typically get 33% of royalties when their images are sold or used."
"Multiple Adobe Stock contributors have expressed concerns over the use of AI-generated images to train their consumer-facing AI tool, Firefly, and that contributors can’t opt out of their work being used to train and create tools like Firefly. Fu confirmed that they are not able to opt out because contributors have signed licensing agreements stating that their images may be used for AI training purposes. But the company says it plans to compensate them in the future when Firefly comes out of beta (Fu declined to say how much it plans to pay the company’s contributors)."
Contributors DID NOT sign any agreement that our work could be use for AI training purposes - that's misleading at best. References to "its treasure trove of high-quality stock images" really irk - Adobe does not own our work. It has the right to issue licenses to customers. Contributors agree to have the images manipulated to operate the site (size and watermarks, for example). The term about using to develop new products is what is purportedly giving Adobe the right to train AI - but that's just Adobe's interpretation of a very vague statement.
Edited to add a link to an article on a boost in Adobe's stock price on the above news, noting they're reporting their earnings next week, on June 15th.
https://www.marketwatch.com/story/firefly-ai-sends-adobes-stock-toward-highest-price-in-more-than-a-year-a92a4709 (https://www.marketwatch.com/story/firefly-ai-sends-adobes-stock-toward-highest-price-in-more-than-a-year-a92a4709)
"Firefly “presents a much needed next act for Adobe” and is a natural extension of the company’s entrenched position within digital media, Turrin said." (Turin is a Wells Fargo analyst)
And yet another article that refers to Firefly training on "Adobe-owned imagery" - that would be our content. Adobe needs its spokespeople to do better in correcting these sorts of mistakes - I know they don't write the articles, but they're trying so hard to pitch the safety of the training data (which I think is much less a slam dunk than they & their lawyers do) that they're not being clear enough that they do not own the content on which Firefly was/is being trained.
https://www.thewrap.com/adobe-firefly-ai-lawsuit-reimbursement-protection/ (https://www.thewrap.com/adobe-firefly-ai-lawsuit-reimbursement-protection/)
Edited Jun 12 to add a link to this article highlighting Adobe's genAI strategy
https://www.barrons.com/articles/adobe-stock-ai-rally-wall-street-ea36f8d2?siteid=yhoof2 (https://www.barrons.com/articles/adobe-stock-ai-rally-wall-street-ea36f8d2?siteid=yhoof2)
"Over the last few months, Adobe has made a series of announcements regarding its growing portfolio of generative-AI software. Until now, Adobe has been giving away its new tools on a website called Firefly, where users can experiment with new photo-editing tools, along with a service that can create images from a simple text command, similar to the Dall-e app from ChatGPT creator OpenAI. But last week Adobe made it clear that Firefly is more than simply a set of digital experiments—Adobe thinks generative AI is going to accelerate its growth rate."
The article said an analyst "...notes Adobe shares in recent weeks have shifted from investors’ AI loser list to the roster of perceived AI winners."
-
What if.. data set usage amounts to something.. someday? If so, what would the AI machine like for breakfast? Are there needed niches that may result in future income?
-
Adobe announced their Q2 results late Thursday and the stock jumped (over $510 in after-hours trading) on the good revenue & earnings growth.
AI was a big element in the positive reaction:
"Analysts have increasingly championed the company since March, when Adobe launched Firefly, a new family of generative AI tools it claims helps transform the emerging technology into something that more design professionals can use, instead of dread."
https://www.marketwatch.com/story/ai-puts-more-buzz-in-adobe-earnings-pushing-shares-over-500-9e8cebb5 (https://www.marketwatch.com/story/ai-puts-more-buzz-in-adobe-earnings-pushing-shares-over-500-9e8cebb5)
The first part of the video is an interview with an analyst who talks about the jury being out on the long-term issues facing Adobe - that if there are fewer creatives because AI does most of the work, who will buy Adobe's creative products - but that in the short term all their AI moves had pleased investors. The analyst also mentioned competition from Canva at the "low end" of the market and getting their Figma acquisition closed (he says they overpaid for it)
https://finance.yahoo.com/video/adobe-scared-bears-away-short-211528215.html (https://finance.yahoo.com/video/adobe-scared-bears-away-short-211528215.html)
https://www.businesswire.com/news/home/20230615449849/en/Adobe-Reports-Record-Revenue-in-Q2-Fiscal-2023 (https://www.businesswire.com/news/home/20230615449849/en/Adobe-Reports-Record-Revenue-in-Q2-Fiscal-2023)
https://www.proactiveinvestors.com/companies/news/1018057/adobe-posts-record-2q-revenue-on-strong-cloud-demand-1018057.html (https://www.proactiveinvestors.com/companies/news/1018057/adobe-posts-record-2q-revenue-on-strong-cloud-demand-1018057.html)
-
Good for them. The stock was clearly undervalued. Adobe is one of the best tech companies that will benefit immensely from the new ai trend.
-
https://www.investors.com/news/technology/adbe-stock-adobe-expands-generative-ai-tool-to-over-100-languages/ (https://www.investors.com/news/technology/adbe-stock-adobe-expands-generative-ai-tool-to-over-100-languages/)
https://www.businesswire.com/news/home/20230712104690/en/Adobe-Firefly-Expands-Globally-Supports-Prompts-in-Over-100-Languages (https://www.businesswire.com/news/home/20230712104690/en/Adobe-Firefly-Expands-Globally-Supports-Prompts-in-Over-100-Languages)
https://techcrunch.com/2023/07/12/adobe-says-firefly-has-now-generated-1b-images-takes-it-global/ (https://techcrunch.com/2023/07/12/adobe-says-firefly-has-now-generated-1b-images-takes-it-global/)
https://www.thurrott.com/cloud/285563/adobe-firefly-now-supports-over-100-languages (https://www.thurrott.com/cloud/285563/adobe-firefly-now-supports-over-100-languages)
Firefly can now be used with more languages, not just English
"In addition to supporting text prompts in scores of new languages, Adobe is adapting the Firefly user interface to over 20 languages. Starting today, it is offering versions of Firefly in French, German, Japanese, Spanish and Portuguese."
"Unlike other players in this field, Adobe can ensure that the images businesses create with Firefly are commercially safe because it’s trained on a corpus of images that are part of Adobe’s stock imagery service. The company even goes as far as indemnifying its enterprise users."
It is increasingly frustrating, to me, that all these expansions of the "beta" of Firefly and the generative fill in Photoshop continue, many months after the announcement in March, and there is nothing - not even a date for a date - on when contributors will have a compensation model for this wholesale use of our content.
Adobe's stock is at $507.58 this afternoon - it closed at $362.88 on March 21st when Firefly was announced. Adobe continues to derive benefit from messages about how their generative AI will be safe for commercial use (because of the training on Adobe Stock content). That's the major reason their stock has risen so much of late.
Edited to add: ADBE closed at $517.28 Jul 13th.
Edited to add: ADBE closed at $532.23 July 18th!
-
Well, if the stock is so high, they can certainly afford to pay out a generous " base training fee" to us all...We are also doing our best to bring them good work to make the Adobe collection more interesting.
Win win for everyone including the producers who are happy Adobe shareholders :)
-
I'm staying waaaay away from all this AI nonsense, but what's the latest on us being compensated for our work being used by Adobe Generative AI tools?
-
I'm staying waaaay away from all this AI nonsense, but what's the latest on us being compensated for our work being used by Adobe Generative AI tools?
I had fussed about that with the Adobe Express (https://www.microstockgroup.com/fotolia-com/adobe-is-now-rolling-out-ai-powered-features-to-adobe-express/) announcement earlier this month. Nothing but crickets from San Jose. I believe the subtlety of Firefly beta being rolled out with Adobe Express - as opposed to Firefly being out of beta - got lost by headline writers. The original promise was that we'd hear about compensation when Firefly was out of beta...
So I think the direct answer to your question is that we know no more than back in March when Adobe announced Firefly.
If I consult my (admittedly broken) crystal ball, I'd say that based on progress so far, Firefly will be in beta for years. Making guesses about Adobe's goals with this announcement, I think it's all about getting the AI buzz wound up for the company as a whole, primarily related to getting the stock price up, and that we (contributors to Adobe Stock) were just the necessary CYA for the messages about Adobe's AI stuff being safe for commercial use. So Firefly could never come to market as a product and Adobe would still be able to win.
There was never anything explicit said about compensation for generative fill in Photoshop even though what I read said to me that it was based on the same training.
Additionally, if you consider that all Adobe's genAI competitors (Midjourney, Dall-E, Stable Diffusion...) are producing the 14+ million genAI collection at Adobe Stock - the stuff that Adobe was contrasting itself with and was painted as questionable for commercial use - the fundamental illogic seems glaring to me. Investors appear not to be paying attention to small details like that.
So my based-on-nothing-but-my-own-flawed-analysis guess is that you shouldn't book a vacation paid for by your Firefly compensation any time soon :)
-
I'm staying waaaay away from all this AI nonsense, but what's the latest on us being compensated for our work being used by Adobe Generative AI tools?
I had fussed about that with the Adobe Express (https://www.microstockgroup.com/fotolia-com/adobe-is-now-rolling-out-ai-powered-features-to-adobe-express/) announcement earlier this month. Nothing but crickets from San Jose. I believe the subtlety of Firefly beta being rolled out with Adobe Express - as opposed to Firefly being out of beta - got lost by headline writers. The original promise was that we'd hear about compensation when Firefly was out of beta...
So I think the direct answer to your question is that we know no more than back in March when Adobe announced Firefly.
If I consult my (admittedly broken) crystal ball, I'd say that based on progress so far, Firefly will be in beta for years. Making guesses about Adobe's goals with this announcement, I think it's all about getting the AI buzz wound up for the company as a whole, primarily related to getting the stock price up, and that we (contributors to Adobe Stock) were just the necessary CYA for the messages about Adobe's AI stuff being safe for commercial use. So Firefly could never come to market as a product and Adobe would still be able to win.
There was never anything explicit said about compensation for generative fill in Photoshop even though what I read said to me that it was based on the same training.
Additionally, if you consider that all Adobe's genAI competitors (Midjourney, Dall-E, Stable Diffusion...) are producing the 14+ million genAI collection at Adobe Stock - the stuff that Adobe was contrasting itself with and was painted as questionable for commercial use - the fundamental illogic seems glaring to me. Investors appear not to be paying attention to small details like that.
So my based-on-nothing-but-my-own-flawed-analysis guess is that you shouldn't book a vacation paid for by your Firefly compensation any time soon :)
And yet people are still using this "beta" feature for commercial purposes. But I guess nobody else but me cares.
-
I just tried it now, but it's so bad. It's not like Midjourney. It's not usable for AI stock photo generation imo yet.
-
We've had a few nights now over a beer and trying to get Firefly to generate anything remotely accurate or lifelike to improve an image. And failed.
Every single thing is unusable from a realism or accuracy point of view.
In one example i selected a group of people i wanted removed from the scene, generative fill selected (no prompts), it removed the people and replaced the background pattern nicely BUT for no clear reason also added a large dog sized mutant pigeon on the pavement.
At no point did i ask for the pigeon version of Godzilla.
As a tangent - the "Remove Tool" in beta. Is that firefly based (i assume not as it actually works)? Or can it be legitimately used instead of content aware to remove items in our own images.
-
Adobe has now announced pricing for Firefly (which is out of beta)
https://techcrunch.com/2023/09/13/adobes-firefly-generative-ai-models-are-now-generally-available-get-pricing-plans/ (https://techcrunch.com/2023/09/13/adobes-firefly-generative-ai-models-are-now-generally-available-get-pricing-plans/)
"Basically, every time you click ‘generate’ to create a Firefly image, you’ll consume one credit (and the company retooled the Firefly web app, for example, so that it doesn’t automatically start generating images before you’ve made all of the tweaks you wanted to make)."
1,000 credits come with the All Apps plan, 500 with a single App subscription. If you run out of credits, things run, but much more slowly.
Press release: https://s23.q4cdn.com/979560357/files/091323AdobeCommercialFirefly.pdf (https://s23.q4cdn.com/979560357/files/091323AdobeCommercialFirefly.pdf)
I understand all the words in this statement, but not what it really means:
"Adobe is introducing Adobe GenStudio, a new solution that combines the best of content ideation, creation, production and activation to revolutionize the Enterprise content supply chain with the power of generative AI. Adobe GenStudio is a comprehensive, integrated solution for any Enterprise to supercharge and simplify their creation-to-activation process with generative AI capabilities and intelligent automation. "
I don't see anything that addresses the AI fueled Photoshop features and the credit system. I assume that means you don't consume credits when you use them, but it'd be nice to understand that better.
More coverage:
https://www.theverge.com/2023/9/13/23871537/adobe-firefly-generative-ai-model-general-availability-launch-date-price (https://www.theverge.com/2023/9/13/23871537/adobe-firefly-generative-ai-model-general-availability-launch-date-price)
https://www.fastcompany.com/90951954/as-firefly-leaves-beta-adobe-promises-bonuses-to-creators-who-trained-its-ai (https://www.fastcompany.com/90951954/as-firefly-leaves-beta-adobe-promises-bonuses-to-creators-who-trained-its-ai)
https://www.zdnet.com/article/adobe-firefly-now-out-of-beta-boasts-fix-for-dall-es-drawbacks/ (https://www.zdnet.com/article/adobe-firefly-now-out-of-beta-boasts-fix-for-dall-es-drawbacks/)
https://www.engadget.com/adobes-firefly-ai-is-now-commercially-available-on-photoshop-illustrator-and-express-130049419.html (https://www.engadget.com/adobes-firefly-ai-is-now-commercially-available-on-photoshop-illustrator-and-express-130049419.html)
https://www.adobe.com/sensei/generative-ai/firefly.html (https://www.adobe.com/sensei/generative-ai/firefly.html)
I haven't seen anything yet about what the rules will be for Adobe Stock contributors who use the AI-fueled Photoshop features to edit a photo they took - does that then become content that must be marked as genAI? With the exception of editorial (which Adobe Stock doesn't except beyond the odd illustrative editorial), one doesn't have to mark photos that have been edited as such. And in general you'd never sell much if you didn't edit what came out of the camera. All these considerations get a lot murkier once there are AI tools to modify a photo of something that actually exists (existed) in the real world.
This is mostly, IMO, a concern about giving the buyer an accurate description of what they're licensing.
-
Adobe has now announced pricing for Firefly (which is out of beta)
I haven't seen anything yet about what the rules will be for Adobe Stock contributors who use the AI-fueled Photoshop features to edit a photo they took - does that then become content that must be marked as genAI? With the exception of editorial (which Adobe Stock doesn't except beyond the odd illustrative editorial), one doesn't have to mark photos that have been edited as such. And in general you'd never sell much if you didn't edit what came out of the camera. All these considerations get a lot murkier once there are AI tools to modify a photo of something that actually exists (existed) in the real world.
This is mostly, IMO, a concern about giving the buyer an accurate description of what they're licensing.
Thanks for the question about how to tag content edited with generative fill, Jo Ann. Here is the official policy which can be viewed in the recently updated generative AI submission guidelines page: https://helpx.adobe.com/stock/contributor/help/generative-ai-content.html
Label your image as generative AI when the use of generative AI tools in Photoshop or Illustrator changed, augmented, or added a new primary subject of an image.
Examples when to label your image as generative AI:
-Generating new additions, such as a new person, animal, or object
-Adding new subjects might compromise the accuracy of an image
-Making significant adjustments or changes to a human subject in an image
-Recoloring the primary subject of an image
Examples when you wouldn’t be obligated to label an image as generative AI:
-Extending background for any reason
-Removing IP or other forms of retouching
-Removing distracting objects or people
-Recoloring the background of an image
thanks again,
Mat Hayward
-
And you are raising CC software prices which were already high, that's really not nice Adobe. :-\
"Starting November 2023, Adobe plans for users to be able to purchase additional Generative Credits through a new subscription plan, starting at $4.99/month for 100 Credits." So will we get paid whenever someone uses our image and know which one is it, like we are paid for normal, paid images? It should be similar, why not? I like bonus I got, but I don't like to be paid in bulk, not knowing how many of my images were used and at which price.
All info is from Adobe blog https://blog.adobe.com/en/publish/2023/09/13/ai-creative-cloud-release-pricing-update?sdid=VG52KCB7&mv=social&mv2=paid-owned&fbclid=IwAR0W4AIxqg-zKvR8XNEQT3r80dW5NGWwxFZgn4DnAjUARChtw_3OLRibYEI (https://blog.adobe.com/en/publish/2023/09/13/ai-creative-cloud-release-pricing-update?sdid=VG52KCB7&mv=social&mv2=paid-owned&fbclid=IwAR0W4AIxqg-zKvR8XNEQT3r80dW5NGWwxFZgn4DnAjUARChtw_3OLRibYEI).
-
And you are raising CC software prices which were already high, that's really not nice Adobe. :-\
"Starting November 2023, Adobe plans for users to be able to purchase additional Generative Credits through a new subscription plan, starting at $4.99/month for 100 Credits." So will we get paid whenever someone uses our image and know which one is it, like we are paid for normal, paid images? It should be similar, why not? I like bonus I got, but I don't like to be paid in bulk, not knowing how many of my images were used and at which price.
All info is from Adobe blog https://blog.adobe.com/en/publish/2023/09/13/ai-creative-cloud-release-pricing-update?sdid=VG52KCB7&mv=social&mv2=paid-owned&fbclid=IwAR0W4AIxqg-zKvR8XNEQT3r80dW5NGWwxFZgn4DnAjUARChtw_3OLRibYEI (https://blog.adobe.com/en/publish/2023/09/13/ai-creative-cloud-release-pricing-update?sdid=VG52KCB7&mv=social&mv2=paid-owned&fbclid=IwAR0W4AIxqg-zKvR8XNEQT3r80dW5NGWwxFZgn4DnAjUARChtw_3OLRibYEI).
Currently there is no contributor compensation for images that are generated and licensed from Firefly. We do not have information on future compensation.
-Mat Hayward
-
...Examples when to label your image as generative AI:
...-Making significant adjustments or changes to a human subject in an image
This article has two images at the beginning which show Photoshop work with generative fill. Would the changes to the human's clothing shown count as "significant"? I realize the decisions will be subjective, but given the recent account blockings, I think you can understand contributors' nervousness.
https://gizmodo.com/adobe-firefly-ai-photoshop-illustrator-free-ai-art-1850833489
I've done a lot of Photoshop editing and re-coloring over the years. If I continue to do it the "old fashioned" way and submit it without tagging as AI, I'd be worried that I'd get flagged as having used AI but not tagged my content as such. Can that dialog with edit history help here?
I've just updated Photoshop to v25 and will experiment a bit.
-
And just a suspicion that agencies want AI images marked, so they won't be used to train AI.
lol - that is "exactly" one of the reasons why they are marked as such... not the only reason, but one of them :)
-
5 dollars gives you 100 credits and 400 images...but unless you have a lot of experience the results will probably be less useful than browsing the adobe stock archive.
This is actually well priced, for most customers a normal stock plan will be a better option and only the more experienced designers will add ai for special cases.
One of the big worries is that ai generation will replace normal stock downloads. But with this price level and the fact you still need a lot of practical experience to prompt well, I am not worried about losing sales.
-
.
-
.
-
Mat,
we need an OPT OUT.
-
What I still don't fully understand is the following. Adobe is marketing Firefly as the safe option to avoid copyright hassles. They even guarantee invincibility. But how does a buyer know on Adobe Stock that they are purchasing an image generated by Firefly and not an image created by, for example, Midjourney?
Will there be a separate category for images created by Firefly? Because if that's the case, the sales of the approximately 15 million GenAI images currently in the database (which are not created by Firefly) may dwindle. Because why would a buyer purchase those, if Adobe implicitly suggests there may be some risk associated with them?
-
What I still don't fully understand is the following. Adobe is marketing Firefly as the safe option to avoid copyright hassles. They even guarantee invincibility. But how does a buyer know on Adobe Stock that they are purchasing an image generated by Firefly and not an image created by, for example, Midjourney?
Will there be a separate category for images created by Firefly? Because if that's the case, the sales of the approximately 15 million GenAI images currently in the database (which are not created by Firefly) may dwindle. Because why would a buyer purchase those, if Adobe implicitly suggests there may be some risk associated with them?
Because Adobe has been promoting Firefly as an "ethical" AI, I was quite surprised when I discovered that Adobe has trained Firefly on other "non-ethical" AI images. So I asked about this on the Discord server and got this response from an Adobe staff member: "Adobe accepts Generative AI images on Stock, and Generative AI images from the Adobe Stock collection are a small part of the Firefly training dataset. Contributors who submit these images must agree to our Contributor Additional Terms of Use, which include the representation that the contributor has the necessary rights to the images, that the images comply with all of the terms, and that these images will meet our guidelines around the use of generative AI tools. . All Stock images come with a representation that they do not infringe the IP of others and generative images are received the same way. If content is provided that violates those representations, we take that content down upon notice."
-
istock put up a notice/reminder that they do not accept images using any generative ai tools including generative fill etc...
This is important to remember as many producers will use ai tools in photoshop in their daily work now.
-
What I still don't fully understand is the following. Adobe is marketing Firefly as the safe option to avoid copyright hassles. They even guarantee invincibility. But how does a buyer know on Adobe Stock that they are purchasing an image generated by Firefly and not an image created by, for example, Midjourney?
Will there be a separate category for images created by Firefly? Because if that's the case, the sales of the approximately 15 million GenAI images currently in the database (which are not created by Firefly) may dwindle. Because why would a buyer purchase those, if Adobe implicitly suggests there may be some risk associated with them?
Because Adobe has been promoting Firefly as an "ethical" AI, I was quite surprised when I discovered that Adobe has trained Firefly on other "non-ethical" AI images. So I asked about this on the Discord server and got this response from an Adobe staff member: "Adobe accepts Generative AI images on Stock, and Generative AI images from the Adobe Stock collection are a small part of the Firefly training dataset. Contributors who submit these images must agree to our Contributor Additional Terms of Use, which include the representation that the contributor has the necessary rights to the images, that the images comply with all of the terms, and that these images will meet our guidelines around the use of generative AI tools. . All Stock images come with a representation that they do not infringe the IP of others and generative images are received the same way. If content is provided that violates those representations, we take that content down upon notice."
Haha, but some of those images clearly violate intellectual property (IP) rights. There are quite a number of AI-generated images featuring a certain fruit logo on the back of laptops, for instance, in the existing database. Adobe certainly places a lot of trust in their contributors. So now, the database will contain a mixture of 'ethical' AI-generated images and non-ethical ones (because they were not generated with Firefly). However, since Firefly is also trained with 'non-ethical' AI-generated images, the database is somewhat contaminated from the start, isn't it?
-
istock put up a notice/reminder that they do not accept images using any generative ai tools including generative fill etc...
Sadly most AI contributors will not care and istock will not be able to tell. I think, at least for now, if you look at an AI image in microstock-suitable enhanced full size, in 90% of all cases you can tell it's AI generated and not a real photo - at least for AI images that are supposed to look like photos. With illustrations it is much more difficult.
But if it is actually a real photo and just a part of it was altered with AI - How is a reviewer supposed to tell whether it was altered with traditional photoshop methods or with Firefly?
But I still appreciate the sentiment. iStock certainly was not one of my favorite agencies due to their extremely small 1cent commissions, but they have gained back some of my respect with their anti-AI stance and lawsuit.
-
I think they are clear in their policy, but they have repeatedly said, they would revisit the issue when firefly is available for commercial use.
At some point I am sure they will accept firefly ai content.
Photoshop sets the standards, a year from now nobody will understand why a normal photoshop tool is not allowed.
getty/alamy/envato are also part of the bria project
https://bria.ai/
Maybe they are waiting until their own ai project is truly usable to generate ai content and will allow both that tool and firefly at the same time.
eta
I registered and tried it. Seems to be a variant of dalle with more than square format and slightly better quality.
Not very impressed and I don‘t think with that level of quality it is suitable for commercial use. Firefly is a lot better.
-
Adobe's business is very lucreative: they feed us
money peanuts (but about 30 or 40 millions dollars for peanuts if paid around $0.10 per image) for stealing us the "food" they gave to their filthy beast. Why didn't they pay before doing it? They had to be able to first analyze the excrement of this beast? Agree or not, everyone must feed the beast! This is how it is, and Adobe calls it “developing generative AI responsibly”, like the ironic words of a dictator.
I am sorry, but I was not happy to receive that money, which I would have liked to refuse to preserve my freedom and image rights, as other more “responsible” stock sites permit us by an opt-out.
But the beast will always be hungry, and they will still use our products, perhaps for free. Products made always made more invisible because they are dissolved in the crowd more and more, and mistreated by search engines. But that will be perfectly well sucked up for use by their deep thing.
They have absolute power, who would stop them in this corrupt world on all levels?
Justice? when the only certainty is lawyers win in any case... the money they get from the people who have been abused.
-
Matt, when accepting the terms of use of the Firefly plug-in in Photoshop, a message appears that commercial use of the beta version is prohibited. Does this mean that the Photoshop plug-in is still in beta or is it safe to upload files to Adobe Stock?
-
Matt, when accepting the terms of use of the Firefly plug-in in Photoshop, a message appears that commercial use of the beta version is prohibited. Does this mean that the Photoshop plug-in is still in beta or is it safe to upload files to Adobe Stock?
Firefly is out of beta. You can use it for commercial purposes. Update your main Photoshop on your computer through the Creative Cloud and you should be good to go.
-Mat
-
Matt, when accepting the terms of use of the Firefly plug-in in Photoshop, a message appears that commercial use of the beta version is prohibited. Does this mean that the Photoshop plug-in is still in beta or is it safe to upload files to Adobe Stock?
Firefly is out of beta. You can use it for commercial purposes. Update your main Photoshop on your computer through the Creative Cloud and you should be good to go.
-Mat
the confusion is what commercial means - we can use the images almost anywhere but can't submit these 'commercial' images to Adobe while we can submit all our other images, including ai generated
-
And you are raising CC software prices which were already high, that's really not nice Adobe. :-\
"Starting November 2023, Adobe plans for users to be able to purchase additional Generative Credits through a new subscription plan, starting at $4.99/month for 100 Credits." So will we get paid whenever someone uses our image and know which one is it, like we are paid for normal, paid images? It should be similar, why not? I like bonus I got, but I don't like to be paid in bulk, not knowing how many of my images were used and at which price.
All info is from Adobe blog https://blog.adobe.com/en/publish/2023/09/13/ai-creative-cloud-release-pricing-update?sdid=VG52KCB7&mv=social&mv2=paid-owned&fbclid=IwAR0W4AIxqg-zKvR8XNEQT3r80dW5NGWwxFZgn4DnAjUARChtw_3OLRibYEI (https://blog.adobe.com/en/publish/2023/09/13/ai-creative-cloud-release-pricing-update?sdid=VG52KCB7&mv=social&mv2=paid-owned&fbclid=IwAR0W4AIxqg-zKvR8XNEQT3r80dW5NGWwxFZgn4DnAjUARChtw_3OLRibYEI).
Currently there is no contributor compensation for images that are generated and licensed from Firefly. We do not have information on future compensation.
-Mat Hayward
Well, that's not great.
I assume a lot of contributors (at least I) would rather get a fractional amount of each firefly generation rather than the upfront hush money.
-
Firefly 2 (beta) is now like a brain that's three times larger!!
"Alexandru Costin, Adobe’s VP for generative AI and Sensei, told me that the new model wasn’t just trained on more recent images from Adobe Stock and other commercially safe sources, but also that it is significantly larger. “Firefly is an ensemble of multiple models and I think we’ve increased their sizes by a factor of three,” he told me. “So it’s like a brain that’s three times larger and that will know how to make these connections and render more beautiful pixels, more beautiful details for the user.” The company also increased the dataset by almost a factor of two, which in turn should give the model a better understanding of what users are asking for."
https://techcrunch.com/2023/10/10/adobe-firefly-can-now-generate-more-realistic-images/ (https://techcrunch.com/2023/10/10/adobe-firefly-can-now-generate-more-realistic-images/)
I didn't do much testing, but with what I did with photos, I can't really see any improvement - and it's not ready for prime time use, IMO. I did not try the match feature - where you upload your own photo to match the style. I'd want to look very carefully at what rights I'm giving Adobe over anything I upload.
I had already decided I need to avoid the generative tools in Photoshop 25 if I want to submit work to stock agencies, but the warning popup in Firefly 2 pointed me at terms and conditions that made it clear no contributor can use these tools if they upload to any agency that trains AI on their own collections.
https://www.adobe.com/legal/licenses-terms/adobe-gen-ai-user-guidelines.html (https://www.adobe.com/legal/licenses-terms/adobe-gen-ai-user-guidelines.html)
1. No AI/ML Training
When using our generative AI features, you agree you will use them only for your creative work product and not to train AI/ML models.
This means you must not, and must not allow third parties to, use any content, data, output or other information received or derived from any generative AI features, including any Firefly outputs, to directly or indirectly create, train, test, or otherwise improve any machine learning algorithms or artificial intelligence systems, including any architectures, models, or weights.
Later on it mentions that they might introduce some beta features that wouldn't be OK for commercial use in any context but is very vague about how you're supposed to know (emphasis mine):
"However, if Adobe designates in the product or elsewhere that a beta version of a generative AI feature cannot be used commercially..."
https://www.engadget.com/adobes-next-gen-firefly-2-offers-vector-graphics-more-control-and-photorealistic-renders-160030349.html (https://www.engadget.com/adobes-next-gen-firefly-2-offers-vector-graphics-more-control-and-photorealistic-renders-160030349.html)
https://www.pcworld.com/article/2100888/photoshops-ai-powers-land-on-chromebooks-as-adobe-debuts-firefly-2.html (https://www.pcworld.com/article/2100888/photoshops-ai-powers-land-on-chromebooks-as-adobe-debuts-firefly-2.html)
This describes the illustrator genAI features (beta):
https://www.theverge.com/2023/10/10/23911114/adobe-max-firefly-generative-ai-model-photoshop-illustrator-express (https://www.theverge.com/2023/10/10/23911114/adobe-max-firefly-generative-ai-model-photoshop-illustrator-express)
There's something here about Adobe's approach to safety and content credentials:
https://blog.adobe.com/en/publish/2023/10/10/adobe-firefly-new-features (https://blog.adobe.com/en/publish/2023/10/10/adobe-firefly-new-features)
The stored thumbnail of a reference image used in genAI will supposedly be used to deal with claims that an inappropriate image was used (and not for training Firefly)
https://www.pcmag.com/news/adobe-unveils-firefly-image-2-ai-at-max-conference (https://www.pcmag.com/news/adobe-unveils-firefly-image-2-ai-at-max-conference)
https://9to5mac.com/2023/10/10/adobe-firefly-ai-model-updates/ (https://9to5mac.com/2023/10/10/adobe-firefly-ai-model-updates/)
https://www.zdnet.com/article/adobe-unveils-three-new-generative-ai-models-including-the-next-generation-of-its-firefly-text-to-image-generator/ (https://www.zdnet.com/article/adobe-unveils-three-new-generative-ai-models-including-the-next-generation-of-its-firefly-text-to-image-generator/)
https://www.forbes.com/sites/johanmoreno/2023/10/10/adobes-all-in-on-ai-debuts-new-firefly-ai-models-for-creatives/ (https://www.forbes.com/sites/johanmoreno/2023/10/10/adobes-all-in-on-ai-debuts-new-firefly-ai-models-for-creatives/)
https://www.adweek.com/programmatic/adobe-firefly-generative-ai-offerings/ (https://www.adweek.com/programmatic/adobe-firefly-generative-ai-offerings/)
https://www.fastcompany.com/90964791/adobe-ai-is-coming-for-your-branding (https://www.fastcompany.com/90964791/adobe-ai-is-coming-for-your-branding)
"Adobe also made improvements to its image generation engine, which is trained on its own stock image library. While the model allows Adobe to avoid messy copyright issues faced by other text-to-image generators trained more broadly, the downside is images look like stock photos. Under the Firefly Image 2 Model, though, generative images look less stock-like thanks to additional inputs including human feedback and the company’s “own private data set,” Costin says."
https://arstechnica.com/information-technology/2023/10/adobes-ai-image-generators-get-beefy-updates-including-vector-graphics/ (https://arstechnica.com/information-technology/2023/10/adobes-ai-image-generators-get-beefy-updates-including-vector-graphics/)
"While Adobe claims that the Firefly Vector Model is "the world’s first generative AI model for vector graphics," a quick Google search disproves this claim, showing earlier vector-based AI models that range from commercial offerings to older experimental research. However, it's safe to say that this is the first vector-generating AI model from a company that is currently considered a major player in the image synthesis space."
-
I saw noticeable improvement, but people still look like illustrations and not photos
-
Firefly 2 (beta) is now like a brain that's three times larger!!
"Alexandru Costin, Adobe’s VP for generative AI and Sensei, told me that the new model wasn’t just trained on more recent images from Adobe Stock and other commercially safe sources, but also that it is significantly larger. “Firefly is an ensemble of multiple models and I think we’ve increased their sizes by a factor of three,” he told me. “So it’s like a brain that’s three times larger and that will know how to make these connections and render more beautiful pixels, more beautiful details for the user.” The company also increased the dataset by almost a factor of two, which in turn should give the model a better understanding of what users are asking for."
https://techcrunch.com/2023/10/10/adobe-firefly-can-now-generate-more-realistic-images/ (https://techcrunch.com/2023/10/10/adobe-firefly-can-now-generate-more-realistic-images/)
I didn't do much testing, but with what I did with photos, I can't really see any improvement - and it's not ready for prime time use, IMO. I did not try the match feature - where you upload your own photo to match the style. I'd want to look very carefully at what rights I'm giving Adobe over anything I upload.
I had already decided I need to avoid the generative tools in Photoshop 25 if I want to submit work to stock agencies, but the warning popup in Firefly 2 pointed me at terms and conditions that made it clear no contributor can use these tools if they upload to any agency that trains AI on their own collections.
https://www.adobe.com/legal/licenses-terms/adobe-gen-ai-user-guidelines.html (https://www.adobe.com/legal/licenses-terms/adobe-gen-ai-user-guidelines.html)
1. No AI/ML Training
When using our generative AI features, you agree you will use them only for your creative work product and not to train AI/ML models.
This means you must not, and must not allow third parties to, use any content, data, output or other information received or derived from any generative AI features, including any Firefly outputs, to directly or indirectly create, train, test, or otherwise improve any machine learning algorithms or artificial intelligence systems, including any architectures, models, or weights.
"
This might be a significant limitation in the ability to use Firefly content for commercial purposes (on other stock sites), and one that Adobe should probably highlight more prominently, in my opinion. Or at least clarify what exactly they mean by it. I've tried my best to read Dreamstime's TOS, but I couldn't find anything resembling AI training. However, who knows what they might do in the near future? What if they change their policies (and update the TOS) and my Firefly-created content is already there? Or what if it stays the same, but the images are scraped by Midjourney to train their model? There's not much I can do about that, right?
In many places (here on the forum, in the Discord channel, on Adobe's own forum), there's a lot of questions about whether you can sell Firefly content. The answer is (in short), "yes, now that it's out of beta, you can, as long as you follow the guidelines." But these guidelines are so strict, especially due to this limitation, that you're almost only "safe" if you sell Firefly content on Adobe Stock. And the review times on Adobe Stock are so long that you don't even need to upload your Halloween or Christmas content by now because you're already too late by then.
And, ironically, the long review times are mainly caused by the flood of AI content that isn't created with Firefly (and therefore, according to Adobe at least, by less "ethical" generators than Firefly, ahem Midjourney, etc.).
-
I recently got an email from AAA with an image that sure as heck looked like it was AI generated.
-
1. No AI/ML Training
When using our generative AI features, you agree you will use them only for your creative work product and not to train AI/ML models.
This means you must not, and must not allow third parties to, use any content, data, output or other information received or derived from any generative AI features, including any Firefly outputs, to directly or indirectly create, train, test, or otherwise improve any machine learning algorithms or artificial intelligence systems, including any architectures, models, or weights.
Later on it mentions that they might introduce some beta features that wouldn't be OK for commercial use in any context but is very vague about how you're supposed to know (emphasis mine):
Thanks for calling this out, Jo Ann. I want to confirm and clarify that you can upload Firefly outputs to Adobe Stock without violating Adobe's generative AI guidelines. The guidelines you quoted are meant to prohibit users from using Firefly outputs for personal or 3rd party AI/ML models.
Regarding your last point about future beta features, you can be confident that if something comes up down the road, I will be here to help explain as best I can.
Thanks again,
Mat Hayward
-
Can we sell firefly output on Dreamstime?
Or use firefly /generative tools on images and then upload to Dreamstime. Or any other agency that takes ai work?
For us firefly and firefly tools are just part of photoshop workflow.
If it is prohibited to upload anything firefly to other agencies, then please I would like to know that.
I will then only upload ai created with other engines like dall e or stable and and postprocess only with my old Photoshop elements.
Things are getting complicated :)
-
I don't know if Dreamstime has said that it will train on its collection, but Getty/iStock have.
I think I can continue to use Photoshop 25.x but will need to stay away from using generative fill to stay within Getty/iStock's rules (weakly enforced at the moment, but...) about not uploading AI content.
I think I might be violating Adobe's rules about training on output from Photoshop's genAI features if I decided to risk it. Getty would be a third party training on content produced with Adobe's features.
As I intend to produce one version of anything I plan to upload to multiple agencies, the only option is to avoid using generative fill & friends. Firefly would get quality rejections so it's not an issue.
-
Can we sell firefly output on Dreamstime?
Or use firefly /generative tools on images and then upload to Dreamstime. Or any other agency that takes ai work?
For us firefly and firefly tools are just part of photoshop workflow.
If it is prohibited to upload anything firefly to other agencies, then please I would like to know that.
I will then only upload ai created with other engines like dall e or stable and and postprocess only with my old Photoshop elements.
Things are getting complicated :)
Like I mentioned above, I haven't found anything in the Dreamstime TOS that indicates AI training. Also, I haven't come across any statements from Dreamstime about it. If I've overlooked something, I'd appreciate it if you could point it out. So far, it seems safe to use Firefly content for Dreamstime. Of course, I don't offer any guarantees :)
Getty says they're engaged in AI training, but they don't accept AI content anyway, so that's not a problem. For Dreamstime, I couldn't find any information about whether they are working on an AI engine. In my post above, I've already mentioned that this brings a significant uncertainty. It might currently be safe (according to Adobe's rules) to upload Firefly content to Dreamstime because it seems (again, without guarantees) that they don't use content for AI training. But what if they change that in the future, and your Firefly content is already there?
By the way, it would be nice if Matt could just provide clarity on this. His response to Jo Ann seems to suggest that it really has to be about training an AI model in isolation (so only with the sole purpose of AI training). However, a straightforward yes or no would be much more appreciated.
-
This is madness! Forget about stock for a second, what if we take regular photos for our commercial clients, that we touch up using Photoshop (for instance by using the Generative Fill or Generative Expand tools). What if at some point our clients decide to use those images to train their own AI/ML. How on earth are we to prevent this? Or even know about it? We definitely need an answer from Mat regarding this topic
-
@Kenneth
Thank you!
Dreamstime is basically the only other place where I sell ai in addition to Adobe. I enjoy that they have near instant acceptance, it gives me a good indication what it looks to customers on an agency, especially if you upload a "connected batch".
-
OK,today I generated hundreds of images with Firefly,then I check them one by one,I will choose only the best ones,and then I have to process them and fix.
Firefly made a good impression on me,but there is still a lot to improve of course,for some types of content the generations are simple and I can get the desired result in a single prompt,for other images it takes many attempts,for others I wasn't successful to create what I wanted.
I see that Firefly is progressing,improving quickly,however I noticed that it still has several difficulties and limitations,for example it is not able to create realistic reflections in water in most cases,it has difficulty with generations,such as ship's mast or soccer nets,soccer field demarcation lines,arms and hands and much more.
For the moment,it is still very difficult to generate an image in which nothing needs to be corrected,however,at the rate at which it is progressing,I am quite sure that in a year at most the image generation in Firefly will have achieved almost perfect results.
-
istock put up a notice/reminder that they do not accept images using any generative ai tools including generative fill etc...
This is important to remember as many producers will use ai tools in photoshop in their daily work now.
yes,but I'm pretty sure that some agencies in the future will have to change their policy on AI content and start accepting it otherwise they will lose customers and money,because there is also work behind the simple prompt to create interesting and attractive AI content,and in my opinion,most customers are not interested in generating but quickly buying the content they need.
-
1. No AI/ML Training
When using our generative AI features, you agree you will use them only for your creative work product and not to train AI/ML models.
This means you must not, and must not allow third parties to, use any content, data, output or other information received or derived from any generative AI features, including any Firefly outputs, to directly or indirectly create, train, test, or otherwise improve any machine learning algorithms or artificial intelligence systems, including any architectures, models, or weights.
Later on it mentions that they might introduce some beta features that wouldn't be OK for commercial use in any context but is very vague about how you're supposed to know (emphasis mine):
Thanks for calling this out, Jo Ann. I want to confirm and clarify that you can upload Firefly outputs to Adobe Stock without violating Adobe's generative AI guidelines. The guidelines you quoted are meant to prohibit users from using Firefly outputs for personal or 3rd party AI/ML models.
Regarding your last point about future beta features, you can be confident that if something comes up down the road, I will be here to help explain as best I can.
Thanks again,
Mat Hayward
@Mat, I am a bit confused. Does this also apply to Firefly 2? Because that's in beta and I thought that whatever is in beta can't be used commercially so uploading to Adobe Stock is only possible when Firefly 2 is out of beta. Just as when Firefly 1 was in beta.
-
It's been confirmed on Discord that version 2 is OK for commercial use.