Wednesday 25 January 2023

beta-testing (10.496)

In an attempt to pre-empt some of the legal and ethical concerns in a field fraught with promise, disappointment and unchecked acceleration that our present concessions to copyright and IP are not fit for, stock picture clearing house Shutterstock has partnered with a suit of text-to-art artificial intelligences to offer clients with existing licensing packages access to a limited dataset used to train the generative AI, promising to compensate artists, photographers, models (contributors) whose work was skimmed, studied. Shutterstock also includes the unenforceable catch-all caveat that generated images must not be used to infringe or misappropriate intellectual property or otherwise generate false, deceptive or misleading imagery—which we can’t whether or not divides the onus and responsibility between publisher and patron.