https://www.newswire.ca/news-releases/shutterstock-expands-partnership-with-openai-signs-new-six-year-agreement-to-provide-high-quality-training-data-872644267.html

The stock's up this afternoon, so investors must feel it's a good deal for them

"OpenAI has secured a license for access to additional Shutterstock training data including Shutterstock's image, video and music libraries and associated metadata"

"Shutterstock's high-quality content library, enriched with vast metadata, leads the industry in size, diversity and annotation�making it unrivaled for training AI capabilities."

https://techcrunch.com/2023/07/11/shutterstock-expands-deal-with-openai-to-build-generative-ai-tools/

"Stock content galleries like Shutterstock and generative AI startups have an uneasy � and sometimes testy � relationship. Generative AI, particularly generative art AI, poses an existential threat to stock galleries, given its ability to create highly customizable stock images on the fly.

Contributors to stock image galleries, meanwhile, including artists and photographers, have protested against generative AI startups for what they see as attempts to profit off their work without providing credit or compensation."

https://www.cnbc.com/2023/07/11/shutterstock-stock-up-on-openai-deal.html

Love the jargon - the shares "popped" today

Edited July 13 to add today's closing stock price - $56.95 -and some comments from The Verge

https://www.theverge.com/2023/7/11/23791528/openai-shutterstock-images-partnership

"Unlike other image-sharing platforms like Getty Images, Shutterstock is fully embracing AI � and all the consequences that may come with it. Artists have expressed concerns about their work getting scraped to train AI models, which Getty Images has addressed by banning AI-generated content from its platform completely. ...

While Shutterstock may see its library grow through its integration with DALL-E, it might not save the platform from the legal gray area surrounding AI-generated content."

Not that it would or could happen, but I wonder how this would all legally play out if all the contributors deleted their ports today. I wouldn't be surprised if SS would keep the data finding some weasel wording in the TOS to keep it.

Not that it would or could happen, but I wonder how this would all legally play out if all the contributors deleted their ports today. I wouldn't be surprised if SS would keep the data finding some weasel wording in the TOS to keep it.

Once the data set has been used ONCE to train an AI it doesn't matter anyways. You can delete your images all you want the AI has already been trained with it.

https://www.fool.com/investing/2023/07/12/does-shutterstocks-new-partnership-with-openai-mak/

"The obvious issue is one of proper compensation. For example, a photographer could put photos or videos online to make money. But generative AI image applications might access these images and add them to its dataset. Users could then create images based on the photographer's copyrighted images and make money from them, but the original photographer wouldn't see a dime. And that's a problem. Shutterstock is a marketplace platform -- it needs both content contributors and paying subscribers."

"The problem is that Shutterstock also risks losing paying subscribers as generative AI grows. People increasingly want the features these other applications provide -- it looks like it could be the future of stock photography. But as already stated, it can't risk upsetting its contributor network in the process. It needs to keep both parties happy, which is why I believe that embracing change through its OpenAI partnership is the way to go."

Edited to add this link with an estimate of what SS has paid out to contributors for data set training - no comment yet from SS

https://petapixel.com/2023/07/12/shutterstock-may-have-paid-out-over-4-million-from-its-ai-contributor-fund/

Well they are training their database with consent from contributors, on the assumption that contributors are abiding by the rules. I won't link to an individual, and this is part of a larger set, but I found this vector of an elephant today.

This is clearly AI created, (which Shutterstock don't allow), and then autotraced as a vector, (which Shutterstock don't allow). On top of this, the keywords were spam, (which Shutterstock don't allow) - light, girl, woman, people, happy, person, sky, silhouette.

If AI is being trained on this two trunked elephant, with two and a half legs and a tusk coming out of his backside, then we are all screwed.

Quote from: KuriousKat on July 13, 2023, 10:53

Well they are training their database with consent from contributors, on the assumption that contributors are abiding by the rules. I won't link to an individual, and this is part of a larger set, but I found this vector of an elephant today.

This is clearly AI created, (which Shutterstock don't allow), and then autotraced as a vector, (which Shutterstock don't allow). On top of this, the keywords were spam, (which Shutterstock don't allow) - light, girl, woman, people, happy, person, sky, silhouette.

If AI is being trained on this two trunked elephant, with two and a half legs and a tusk coming out of his backside, then we are all screwed.

or maybe the people depending on AI are screwed

Agreed, but I meant that the likes of Shutterstock and Adobe don't seem to care what's getting into the database, either by fair means or foul.