Sora: Why everyone is talking about OpenAI’s new project

Sora
Why everyone is talking about OpenAI’s new project

The The “Text to Video” program Sora from the company OpenAI is not yet freely accessible, but it has dominated the headlines since its presentation a few days ago. With simple prompts, you can generate photorealistic videos that will excite and worry many people at the same time. One example of many is Hollywood producer Tyler Perry (64), who found Sora’s presentation so impressive that he has put the $800 million expansion of his studio on hold indefinitely. Why Sora is currently making such a name for himself.

There is expressly no schedule for the release of Sora, the AI ​​company OpenAI announced when presenting its new star. People seem to be aware of the potential of video-generating AI – and therefore also its dangers. Because Sora is capable of doing this simple requests such as “The view of an ant on the way through its nest” to generate videos that are deceptively real and can no longer be recognized as an AI product at first glance.

Without even having to get up from your computer, you can create videos of anything we can imagine and put into words. The creative examples from Sora’s beta testers are already fascinating, and as always with these technologies, it’s important to remember that the current state is the worst the technology will ever be in. For many people, however, this very prospect is worrying. How can you protect yourself from the dangers of technology when deepfakes are already a real problem?

Hollywood star sounds the alarm: “No rules, no survival”

Sora’s videos are played with a watermark, but this can easily be cut out. There are therefore calls for OpenAI not to make Sora available to the public until the entire tech industry has agreed on a standard for a digital watermark that unambiguously identifies AI videos as such. Other voices believe that the only way to deal with the upcoming flood of AI content is to train more AIs to recognize generated videos as such.

Sora’s potential is so great that entire industries are paying attention. Who needs providers of stock material in the future when you can simply dictate the desired images? What will happen to the profession of cameramen, editors, make-up artists or actors if an AI can simply imitate them out of ones and zeros? What does such a technology mean for the copyrights of the videos used to train the AI? Hollywood producer Perry says in Interview with “The Hollywood Reporter”: “We need regulations, otherwise I don’t see how we can survive.”

Is Sora part of a much more capable AI?

Sora is also currently the cause of speculation that OpenAI may have already created general artificial intelligence (AGI). Sora would therefore only be a small part of a much more capable AI. In user forums and on tech blogs, for example, it is discussed that Sora must have a sound understanding of physical relationships in order to be able to display the videos with the consistency shown. Technically, there must be more to this than OpenAI has so far communicated. In addition, OpenAI has with “Jukebox” The next project in the pipeline is already one that can compose music and generate sounds, which is another indication of the existence of an even more capable AI.

So far, however, these are speculations that OpenAI CEO Sam Altman (38) has neither confirmed nor denied publicly. Quite firmly denied he on the weekend however Representation of the “Wall Street Journal”, he wants to collect up to seven trillion dollars in funds to build chip factories and has presented plans for this in the Arab Emirates. At an event hosted by chip manufacturer Intel, Altman denied such a plan. It is true that a lot of money needs to be invested in order to bring technology to people that will change the future for the better.

SpotOnNews