Assuming things are correctly labeled - which is a huge assumption when it comes to stock libraries.

On Adobe Stock, if you use the filter "Exclude Generative AI" and then use the search term "generative" there are somewhere around 600,000 images whose titles say they're created with Generative AI, but which aren't tagged and thus will show up in searches where buyers have explicitly said they don't want AI-created work.

Some images may make the mis-labeling obvious though

...but what exactly is wrong with the Roti?

The back of the plate is completely missing - if you look at the PNG view on Adobe Stock it's easier to see the "hole"

And it's not just nonsense lamps, look at the legs of the chairs - one has only three legs and the lengths are such that you couldn't possibly sit in them

And have a "burger with egg" for breakfast while you're here

As I've said a number of times, it is the agencies' responsibility to review content, for the protection of their customers (even if they don't give a toss about us as contributors). This is especially true with new portfolios - lots of the ones you posted had less than 100 in their portfolios.

I understand the agencies are trying to automate and cut the costs of the inspection process, but they're not doing a good job of that. They're fouling their own nests with this sort of short sighted cost cutting.

This struck me as funny (in a gallows humor sort of way). One of the people who stole your mushroom shot had apparently stolen other work and decided to add text to it (genius disguise, that...). Small problem in that he's too lazy to even add the greeting correctly.

Now, the rejections on Adobe get ridiculous. I had two photos of Cartagena (townhall and Roman theatre) rejected because they are too similar to previously submitted photos in my portfolio. I have never been to Cartagena before, so these two are my first photos of Cartagena ever submitted. In my portfolio of 2500 images I have 8 pics of town halls of several cities. Does this rejection reason mean that the townhall of city A is too similar to the townhall of city B? And I am not allowed to submit photos of townhalls of different cities? What worries me, people get their account closed because of similarity. So, when I submit a further townhall from another city I will risk to get my account closed? I don�t understand these rejection rules and how a �reviewer� (AI?) can think the townhall of city A is the same as of city B.

On the "too similar" rejection, that doesn't appear to apply to genAI uploads. I just counted 31 closeup pictures of one eye of a tiger's face - from one contributor and all consecutive (meaning I think they were all uploaded together).  Edited to add 53 near identical "exploding" hamburgers, likewise from just one contributor.

This drunk-on-the-greatness-of-genAI just results in a pollution of an otherwise great collection of content...

If you do a search on Adobe Stock for "generative" and select the filter to exclude Generative AI, you have 644,228 results. I spot checked a number of them and they (a) look like AI and (b) the title quite clearly says they were generative AI (obviously there will be some that have that search term for another reason, but it's still a large number of improperly identified images).

So these predated the filter, or someone forgot to check the box or something - and reviewers didn't notice or don't have a mandate to get that sort of thing fixed?

This isn't just stuff from a long time ago. The newest image in the collection at time of writing was 611697106 and the newest AI image that didn't have the GenAI tag set was 610830747

This must be corrected. Even Adobe's corporate spin at various public events has included the importance of being transparent - tagging AI generated images as such - and surely that has to include the Adobe Stock collection.

"As a trusted partner to individuals and businesses of all sizes, Adobe develops and deploys all AI capabilities with a customer-centric approach and according to its AI Ethics principles to ensure content and data transparency. Content Credentials provide �nutrition labels� for digital content and are a key pillar of Adobe�s AI principles."

https://news.adobe.com/news/news-details/2023/Adobe-Unveils-Future-of-Creative-Cloud-with-Generative-AI-as-a-Creative-Co-Pilot-in-Photoshop-default.aspx/default.aspx

https://contentauthenticity.org/

https://venturebeat.com/ai/adobe-commits-to-transparency-in-use-of-generative-ai/

"And how will we know whether something we see was created by a human or a computer?"

On Adobe Stock, one would hope you'd know because content was properly handled by the upload and inspection process, and ideally, thumbnails would be overlaid by a "Gen AI" box (like editorial and premium images have)

I completely get that "Quality" is a very subjective measure, and that all the agencies have the right to set their own standards for what they will and won't accept, but as I look through the 10+ million (acknowledged) genAI images in the Adobe Stock collection, I think there's work that should be rejected because it's faulty reality - not artistic point of view, not fantasy, not creative expression, but just a mistake.

I've been collecting examples - roast turkeys with four drumsticks and two wings; a lobster with three claws (the one in the front sort of circular); an office chair with three arms, one in the middle so you can't sit; a cow with a single bump-like udder in the middle of its belly and another with several balloon-like udders in the right (ish) place; an office chair with two legs and one arm that couldn't even stand up; a refrigerator with open doors too wide to close; a photo of London in the 1940s which not only isn't but is mismatched building bits, US flag with too many/few stars, stripes, stars that are squares or triangles, spiders with 10, 12, 14 legs and cockroaches with 8 . . .

Two issues with this type of thing from my perspective. One is that this type of content is defective - noise free, in focus, good color, but just a mistake.

The other is that some buyers may not realize the content is erroneous and just assume the agency has their back and it's good to use. An example: a buyer has not been lucky enough to eat lobster or even know what it looks like, but has been told to download a picture of a lobster meal. GenAI content is included in search results by default - you have to turn it off if you don't want it, so the unlucky buyer downloads a mutant genAI image with extra claws without realizing it's pretty but wrong. Doesn't an agency have some responsibility to weed out this type of "mistaken" content?

EDIT to add: I mean what lol:
https://stock.adobe.com/in/images/health-and-healthy-lifestyle-concept-man-riding-a-bicycle-with-a-bunch-of-fruit-generative-ai/599130317?

That now generates a 404 error - perhaps the crack inspection team was embarrassed at what they had allowed into the collection now someone pointed it out?

Don't you need to pay commission from sales to Midjourney? I remember saw it in their licencing.

In order to "own" your output from Midjourney, you have to have a paid subscription plan - otherwise they own it. "If You are not a Paid Member, You don�t own the Assets You create. Instead, Midjourney grants You a license to the Assets under the Creative Commons Noncommercial 4.0 Attribution International License (the �Asset License�)."

You can see the plans they offer here:

https://docs.midjourney.com/docs/plans

"If you have subscribed at any point, you are free to use your images in just about any way you want. If you are a company making more than $1,000,000 USD in gross revenue per year, you must purchase the Pro plan. For complete details please see the Terms of Service"

As noted above, I don't see anything in the terms of service about paying Midjourney, other than the subscription. They do note that if you're sued you're on your own (their words are different, but that's the gist of it)

I took a look with the display zoomed - it's really hard to evaluate such small thumbnails - but here are my guesses:

1- AI - everything super-green super-lush & the cows aren't fenced in
2- AI - barren valley & lush hilltops with chocolate-box art vibe
3- AI - lighting looks fake
4- AI - shoulder blade looks unreal and the glass stem is unusually long
5- AI - Lights softly hazy and shadow areas very pale; greenery blobby & uniform color
6- Reality Way over-processed, but looks like a cleaned up long exposure shot
7- Reality
8- AI - I don't think there's a view like that or columns/tiles like that in the real thing
9- AI - Odd combination of rounded boulders and super-jagged peaks and grasses up that high?
10- AI - lighthouse looks tiny and unclear where the light on the water is coming from with such dark clouds
11- AI - hands
12- AI - lighting on ball doesn't match lighting on dog (but it could be a bad composite)
13- Reality
14- Reality
15- Reality
16- AI- lighting and doll-like quality of child
17- AI - composition & lighting
18- AI - dystopian scene & lighting
19- AI - it's all too perfect - none of the stock photos of scallops look this good
20- AI - it's all too perfect - none of the stock photos of scallops look this good
21- Reality - focus isn't good, but damaged wood is such a favorite subject...

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.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

"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/

"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

"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

"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/

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

"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."

https://www.prnewswire.com/news-releases/shutterstock-announces-share-repurchase-program-301844432.html

Unsurprisingly, SSTK is up this morning - it had been wallowing around in the high $40s recently and it's now at $52.26

From the CEO:

"Given our strong free cash flow generation and healthy balance sheet, Shutterstock is uniquely positioned as a technology company to be able to invest for organic and inorganic growth while also consistently returning value to shareholders through a mix of dividends and share buybacks,"

They've authorized up to $100million, and apparently can borrow to fund this cozy-up-to-the-shareholders scheme:

"The Company expects to fund repurchases through a combination of cash on hand, cash generated by operations and future financing transactions."

I guess Oringer would love to get more for his program of stock sales he still has 11,370,600 left. June 2nd he sold nearly $4million and the price in the low $50s was way down from the April sale in the $70s

https://finance.yahoo.com/screener/insider/ORINGER%20JONATHAN

Having squeezed contributors as tightly as they dare and banged the AI drum as loudly as they could the stock price wasn't going up as they'd hoped...

Tossers!

Edited to add: The stock closed today (Wed June 7th) at $50.63, up only a few cents from yesterday's $49.99 close; it's $50.20 in after hours trading. I guess the buyback exuberance  fizzled.
June 8th close $49.80 (and the markets were up; this was about SSTK specifically, not general doom & gloom)
June 16th close $49.13 (and Oringer sold nearly $5million more stock on June 7th at $50-$53 per share)

I am getting a lot more photo rejections than usual. I keep reading that people use topaz to denoise files especially for Adobe...

I don't think noise is the issue (although obviously I haven't seen other people's photos except for a few posted here). Firn's rejection doesn't look like a quality issue to me either.

I've lost patience with the randomness of what's accepted at Adobe Stock and what's not - and with the total lack of information about why.

I'm going to take an uploading break for a bit. Life's too short...

SS added a blog post summarizing the event

https://www.shutterstock.com/blog/ai-showcase-highlights-and-actionable-tips

I received email today asking me to take a survey: "Help transform our innovative generative AI concept into a practical enterprise tool". I took the survey, although I'm clearly not their target customer. I'm not sure how useful anyone's answers will be when some of the setup text was so vague.

For example: "Consider the following offerings that use generative AI to automate and optimize custom content creation, resulting in unique and engaging content and solutions. Which of these would you or your company be interested in using?" It's all about the end result - unique and engaging doesn't really tell me whether I'd like it.

They asked questions about how I'd like to access and pay for AI tools and content - pay when I license and download; pay a fixed subscription for so many assets per month; a credit system; an unlimited use subscription.