Stock images are used for many purposes, but finding the right one can be hard, especially if we consider both motive and quality. That is why Getty Images and Nvidia have launched Generative AI by iStock, a text-to-image platform that can create stock photos from text descriptions.
Generative AI by iStock is a web-based tool that uses Nvidia’s Picasso model, a type of artificial intelligence that can generate new images by learning from a large dataset of existing images. Generative AI by iStock only learned from Getty’s creative library and iStock’s stock photo library, which contains millions of high-quality images.
Generative AI by iStock can help users get the images they need, which can make some impossible images viable and even easy, but can also send the “ancient” photographers who still produce high-quality stock images even closer to extinction, as the same AI can easily generate similar content without the required investments from both the photographer and buyer.
Getty will offer legal indemnity for the images its tool generates, with a cap of $10,000 per asset. The company will also offer contributors whose content was used to train the model the possibility to participate in a revenue-sharing program.
Generative AI by iStock is a new way to create stock images, but it is not the only way. It is a part of the larger ecosystem of stock images, which includes human photographers, artists, editors, and consumers. It is a new opportunity with its limitations and challenges, such as the quality, authenticity, and diversity of the images it produces.
Creators and buyers have the responsibility to use it wisely and ethically, but they also have the power to send stock photographers to the history books. Will AI-generated images ever achieve the same quality? And most important, will buyers give quality enough weight or will they settle for not as good but much cheaper? At the moment, there are lots of considerations regarding the quality of the images GenAI can create but the gap is closing, and very fast. Months ago it was very hard to generate text, hands would always look alien, and expressions and emotions were not even considered. Fast forward just a little bit and most of those issues are either fixed or on the way to it. As AI-generated images improve, they bring down the barriers preventing more people from joining the AI train. The easier it gets, the less gear is needed to create images, and more “less techy people” can join. This shift in the ever-more-open industry will certainly kill some jobs as they were in the (almost dead) past, but it will also shoot up exponentially the opportunities for creators in a more democratic environment.
What do YOU believe the future holds for stock photography and photography in general?
I leave you to think about it while enjoying a penguin family holiday stock photo album: