The leak was auf Hugging Facean open source community platform for machine learning, shared by people who said they had access to the tool as testers. The model is said to be able to create 10-second videos with a resolution of up to 1080p.
Unpaid work for a multi-billion dollar company
Under the username PR Puppets, the people involved in the leak accuse the company of exploiting labor. The approximately 300 testers, including artists and filmmakers, provided OpenAI with valuable feedback and creative input, but without being compensated appropriately or at all. This is what the post says:
Hundreds of artists provide unpaid labor through bug testing, feedback and experimental work for the program for a $150B valued company. While hundreds contribute for free, a select few will be chosen through a competition to have their Sora-created films screened — offering minimal compensation which pales in comparison to the substantial PR and marketing value OpenAI receives.
Ethical questions reignited
The leak raises long-discussed questions regarding ethical standards and copyrights in the context of image and video development with AI. The artists involved in the leak particularly complain that early access served less for creative expression and was more aimed at using the testers – sometimes unpaid – for advertising and PR activities.
As early as April 2024, OpenAI faced harsh criticism when it became known that the company had used publicly available YouTube videos as training material for Sora. YouTube CEO Neal Mohan then emphasized the problems of such an approach. How OpenAI itself will ultimately react to the current leak remains to be seen. However, the company has now taken action and temporarily blocked access to Sora for all artists.
Source: onlinemarketing.de