The plaintiff artists provided illegal access to Sora’s artificial intelligence • DigiKala Meg

Openai, a pioneer in the field of artificial intelligence, unveiled Soras last winter, a smart text -to -video conversion model. Later, the company allowed a limited number of users to test the model. Yesterday it was reported that a group of these experiments, in a protest move, revealed the technology to the public.

Recently, a group of people who had access to the Sora Video Video Model have said they have disclosed access to the public. On the Hugging Face platform, they launched a project that was apparently connected to the Soras API. The project operated on the authentication tokens that probably previously provided to test this artificial intelligence, allowing users to produce videos.

The project was deactivated shortly after its release, but a number of users on the X -Social Network have announced that they have been able to use it, and its outputs have had well -known features of Openai products. The disclosure group claims that after three hours, Openai discontinued all artists from accessing the Sora model.

But what caused this revelation? The group says Openai had called on the early Sora experiments to provide a positive picture of the model, but it had not appreciated their efforts. A part of a note published by the group reads: “Hundreds of artists trying to find bugs, providing feedback and testing without receiving wages for the Soras initial access program, which belongs to a $ 150 billion company. They have. Apparently, this program was more for advertising and public relations than a creative expression and critical evaluation. “

The revelations claim that Openai has given people the wrong information about the capabilities of Sora and has tried to hide the reality by keeping access to this model. All videos produced by this model must be approved by Openai before being released, and the company only examines the outputs of a small group of experiments. In response, Openai said in a statement that the project was voluntary and that participants had no mandatory to provide feedback or use of the tool.

Source: Techcrunch