Screenshot of OpenAI's official website.
Just a few words can generate an explosion of 60 seconds**! SORA, a generative model launched by the Open Artificial Intelligence Research Center (OPENAI) in the United States, has shocked the world recently.
It's only been more than a year since OpenAI launched ChatGPT, and it has dropped a bombshell again. What is the amazing thing about the "Wensheng**" AI model SORA? What is the impact on the industry? What are the hidden dangers? Xinhua News Agency reporters will solve the situation for you
What is Sora capable of?
A number of short ** "blockbusters" generated by SORA are swiped on the whole network, and the scenes are realistic and silky, and the details are rich.
This is OpenAI's first foray into the field of AI generation. According to the company, Sora uses the Transformer architecture to create realistic and imaginative scenes based on text commands, generating high-definition ** in a variety of styles, in different formats, and up to one minute.
Sora is also capable of generating from a still image, expanding an existing one, or filling in missing frames.
OpenAI says that Sora understands language deeply, not only understanding user text prompts, but also understanding how said things exist in the physical world. "We are teaching AI to understand and simulate the physical world in motion, with the goal of training models to help people solve problems that require interaction with the real world. ”
However, is the ** work posted on OpenAI's official website the average level or the highest level created by Sora?
OpenAI admits that the current SORA generation** may contain illogical images, confuse spatial details such as left and right, and make it difficult to accurately simulate the physical principles and causal examples of complex scenes. For example, a person may have taken a bite of a cookie, but there are no bite marks on the cookie. However, with the enhancement of computing power and the improvement of models, people may get more complete and advanced generation functions in the short term.
Whose jobs may be smashed.
OpenAI's launch of SORA this time is more like a preview version, and it is not yet easy for the public to fully understand the advantages and disadvantages of the model. OpenAI says that at the moment it is only providing SORA access primarily to a select group of people, such as designers and filmmakers, to get feedback on improving the model. Not only did the company not release the underlying details such as the data on which the SORA model was trained, but it also did not determine when it would be released to the public.
Sora generates a screenshot.
Some analysts believe that SORA has once again highlighted the far-reaching impact of AI advances on real life and traditional industries. The huge development prospect of artificial intelligence in the field of generation will open the door to shape the new format of the film and television industry, and at the same time, it may subvert the existing film and television industry.
On the second day of SORA's launch, the stock price of Oldobie, which is mainly engaged in image processing and production software, exceeded 7%.
Hollywood suffered its first industry-wide strike of writers and actors last year in 63 years, as some jobs in the industry could be replaced by artificial intelligence. The emergence of SORA has made this threat even more imminent and real.
Make forgeries more realistic and difficult to identify.
OpenAI wrote in the title of SORA's technical report, "The generative model is a "world simulator."
If the world can be simulated, what is the boundary between truth and falsehood? Many in the industry are worried that SORA will fuel the "deepfake" technology. "When news, images, audio, anything can be faked, then in that world, nothing is real," said Farid, associate dean for the School of Information at the University of California, Berkeley. ”
On November 2, 2023, at Bletchley Park, England, a participant walks past the promotional display board of the first AI Security Summit. Photo by Xinhua News Agency reporter Li Ying.
In response to the concerns about counterfeiting, OpenAI said that when it actually launches its products to the public, it will ensure that metadata is generated and tools to detect authenticity. OpenAI also pledged to take security measures before using SORA in its products, including adversarial testing of the model by "experts in areas such as misinformation, hateful content, and bias" to assess harm or risk; Verify and reject text input prompts that contain extreme violence, sexual content, hateful images, other people's IPs, etc.
However, OpenAI admits that even with extensive research and testing, "we can't do all the beneficial ways that people use our technology and all the ways in which we misuse it."
Can AI Hurricane manage it?
Disruptive innovation in the field of science and technology is constantly emerging, and how to achieve a balance between embracing technological progress and ensuring social security has attracted more and more attention from all walks of life.
OpenAI said it will work with policymakers, educators, and artists around the world to understand their concerns, identify positive use cases for SORA, and believe that learning from real-world use is a key component in creating and releasing increasingly secure AI systems.
On July 7, 2023, at the "Artificial Intelligence for Good Global Summit" in Geneva, Switzerland, a participant interacts with the ** robot Sophia. Photo by Xinhua News Agency reporter Lian Yi.
Industry insiders pointed out that in the current situation where the current governance framework and control measures have not kept up, enterprises alone may not be able to provide the AI security and trust required by society.
Editor-in-charge: Ren Haopeng |Review: Li Zhen |Supervisor: Wan Junwei.
*: Xinhua News Agency).