Shutterstock has marked a historic agreement with OpenAithe company behind the now famous From E 2, to start selling stock-like images generated by artificial intelligence. According to what was anticipated, there will be direct access to the performing tool from the Shutterstock web portal, which has promised to protect the authors of the photos that were used to train the i.a. with a percentage of revenues, as a reward for contributing to the development of technology. The other colossus of the photos remains on their positions, Getty Imageswho through his CEO commented that it could be a move beyond the limit of the regulation.

Stock photo portals seem like one of the perfect applications for creations generated by artificial intelligence, a tool with almost unlimited potential. Customers rely on these immense databases to search for a specific image to accompany, for example, an article or an advertising campaign and use a textual description as a starting point for the search. In short, the same procedure as text-to-image already seen on Dall-E 2 and associates, for this reason Shutterstock was quick to make a deal which combines the world of photos taken by a camera in the real world and those born from an i.a. The goal is not only to monetize, but also to reward the authors of the images used for training of the system, which often feel deprived of their digital properties.

See also  Did Spotify screw up? ‘No… and yes,’ says its CEO

The service will become operational in the coming months and Shutterstock will ban all those images made by other AI. And while Google keeps its Imagen private until there is a way to exploit it in a completely legitimate way, Getty Images CEO Craig Peters, he expressed himself on the rivals’ move confirming the total ban on the generators and stating that there are still too many details to be clarified and that such a casual approach could be “Dangerous, not responsible and potentially illegal“. There are methods to check if your photos have been (unduly) used to train artificial intelligences, but they don’t cover all databases, so the gray area is quite large.

.


Source link

Leave a Reply

Your email address will not be published. Required fields are marked *