Before the Spring Festival, the technology community ushered in three major news.
First, Nvidia released "chatwithrtx", which can make everyone's computer a large operating system for localization;
The second is that Google AI has released its GEMNI15. Compared to GPT-4, it can handle more than 1 million tokens. These two events seem to be out of step with ordinary people's lives.
But the next piece of news exploded in everyone's circle of friends. This is a new generative AI model released by OpenAI.
NetnetworkscaledworldwillhappensGiantchange
On the morning of February 16th, OpenAI officially released its first text generation mode, SORA. Users can simply enter text commands and Sora will instantly generate a 60-second video with detailed backgrounds, themes, flexible multi-view shots, and rich characters. 
Two days after Sora's launch, it became a hot topic on the internet, with speculation that it would revolutionize the film and television industry.
According to the MIT Technology Review, SORA is not intended to be used by the general public, and only relevant software developers, security testers, and a small number of creators and artists selected by OpenAI can use it.
In addition to dealing with regulations and the harm it can bring, OpenAI's first task is to put SORA in the hands of visual artists, designers, and producers. It can be seen that this series of literary creation models will be oriented to films, films and other industries, and at the same time, open AI also hopes to get corresponding feedback from expert evaluations, so as to promote the development of the model.
Prior to this, generative AI is gradually reshaping the operating model of media, advertising, finance, education and other industries through automation, personalization, optimization and other means, improving production efficiency and optimizing the decision-making process. The emergence of SORA heralds the arrival of AIGC (AI), and tools like SORA, which will have a profound impact on future content and business models. 
We found that OpenAI recently released 48 demos made by Sora, and we have conducted many research and analysis on this and have come to the following conclusions:
SulaCauseAttackgrams"Physical puzzles".WhileStand out from the crowd
Compared to the 10-second creation time of other products, the ** time generated by ora is 60 seconds, and the details of the produced content are so delicate that there is almost no difference from the real world. Also, the consistency between the theme and the environment in the content produced by SORA is amazing.
But most importantly, Sora's ability to model the world. During this time, Sarah learned a lot about the real world and the laws of physics. However, existing algorithms can only convert text and 3D models normally, but cannot directly embed them into the real physical world.
In short, it is the literary works of the past, which give people a feeling of "moving **", because of the lack of the concept of "real", therefore, it cannot meet the needs of users.
Taking Sora's "A Pirate Ship Fight on a Coffee Cup" as an example, the physical problems such as surface hydrodynamics and action scale are studied.
Jim Fan, a senior R&D officer at NVIDIA, even said: "SORA is a data-based entity engine" and "it is a first-class software that can be learned, or a "global model".
2. sorawill be usedAmazingofSpeedUpdate
Similar to the previous ChatGPT, SORA is a big data-based AIGC tool that has similar network effects to ChatGPT, with a very fast iteration process and is not calculated according to Moore's Law.
However, unlike Chat GPT, the AI content generated by SORA is more convincing for the average person and easier to get feedback from users and practitioners. As the main form of entertainment and information medium in today's society, its rich information resources and communication channels provide strong support for SORA's independent learning and improvement.
Although Sora also had misconceptions such as "juice in a cup will spill" in the early days, it also shows that Sora has been independently exploring the material world, and this method of self-correction and self-learning is more suitable for the definition of "intelligence" by human beings.
Therefore, Sora's understanding of the work will become higher and higher, and the overall has far surpassed the previous most popular Runway and Pika, with the accumulation of massive data, the future of Sora has not yet come to an end.
Zhou Hongyi, the founder of 360, said, "The AI of open AI will watch all the movies and **, and it will not be too long for AGI to be completed by then, maybe one to two years." ”
GeneralHumansoraonWriteCome outofstuffNot reallythanMostlyofSpecializedHomePoor
In the past, to buy a high-cost camera, you had to learn to adjust various parameters, and to wash out a photo, you had to use complex chemicals and go through many processes. However, with digital cameras and smartphones, ordinary people can take photos and view them anytime and anywhere. As a result, the traditional image of a professional photographer can gradually only be expressed through some abstract elements such as composition, lighting, and personal style.
Now, Sora is doing the same thing. Coupled with its huge model and ever-improving self-learning ability, SORA can make a person who knows nothing about movies, art, art, etc., make a matching image according to their imagination in a short time, and its effect is not inferior to those high-budget sci-fi blockbusters.
This also means that in the future, in addition to writing their own articles, online writers can also independently complete a fantasy short drama by themselves. Especially for those who specialize in special effects and virtual sets, it is even more stressful to spend a lot of money to make something that is similar to what those grassroots writers wrote.
TrueofAbundanceRichofEmotionContentsRightShortCreatorsofValueYesNoneStopof
Driven by tools such as SORA, with AIGC being promoted by tools such as SORA, short ** has become the main way to obtain information, and there will inevitably be a large amount of AI content, while real scenes and narratives are the "hardest hit areas".
Moreover, with the development of AIGC technology, it has become difficult for users to distinguish which ones are created by AI, so the distinction between users and creators has become smaller and smaller.
However, no matter how deep Sora's "understanding" of the world is, no matter how vivid it is, it is only a digital product that can never replace the real image. Therefore, if you want not to be overwhelmed by mass-produced ** in the future, you can only get better results by focusing on reality and impressing the audience with deep emotional expression.
Sulaof
Ever since the launch of SORA, the film and television industry has been the talk of the talk. Among them, SORA produced the best results, which are not inferior to Hollywood blockbusters. Many people believe that in the fields of film, television, and film, SORA will greatly reduce the production mode and industrial chain of film, television, film, television and other industries.
At the same time, our group also shared some short films produced by SORA with colleagues in the film and photography industry. Some visual professionals believe that in the ** of realistic subjects, the effect produced by SORA is not realistic enough. When ordinary people see this scene, they feel a little weird at most.
They believe that Sora's CG (Computer Animation) is doing so well that even the works of other CG companies can't compare to it.
While Sora is gifted at making CGs and can make 60-second short films for very little money, a film doesn't consist of 60-second short films.
Movies and TV series often require big scenes, big scenes, but SORA is not currently able to meet this requirement, and the content produced by SORA is difficult to withstand professional scrutiny in terms of details.
A 60-second short film is fundamentally different from a feature-length film in terms of narrative style, and similarly, AI can write good jokes, but it is difficult to write a good work of more than a million words.
We found that while SORA has a strong ability to generate and learn, it still can't meet the demand for massive amounts of information and detailed information. Now, in film and television production, humans cannot be completely replaced. However, there is no doubt that Sora is an excellent tool for creating concepts. In particular, its unique advantages in small scenes** will have a profound impact on the advertising industry and the creative industry.
Another attraction is the platform. According to statistics, China's short-term ** users have exceeded one billion, and Douyin's revenue has exceeded $70 billion in 2022, and Kuaishou's annual revenue has reached 90 billion.
However, in the face of this huge market, any of the best production tools have not yet been commercialized or industrialized until SORA is launched.
Compared with Pika and Runway, Sora has a better performance in picture quality, and it is more than 10 times longer than **. If it is applied to short** creation, it can greatly improve the freshness of the user. However, when many creators are using SORA for "homogenization", how to ensure the excellent quality of their products has become the biggest problem for SORA.
The rise of AIGC has made "imposition", "following the trend" and "short**" a new trend, making it impossible for creators who rely purely on original content to survive in the competition. Judging from the current situation, self-**, which is dominated by ** and narrative self, may be able to make better use of the advantages of SORA.
Therefore, for the creators of the ** platform, SORA cannot replace their creativity and inspiration, but is only an auxiliary tool to improve the efficiency and quality of creation.
From Sora to now, the best works are focused on animals, vistas and fantasy scenes. In the past, such films often cost a lot of royalties to be used. Therefore, the emergence of ORA may be the first to break the material copyright industry.
Because SORA has a strong self-replication ability, it is extremely destructive to social order, so it is also easy to be used by people with ulterior motives to carry out criminal activities such as fraud, extortion, and defamation.
It can be expected that SORA will face strict monitoring in many countries and regions around the world before it is officially launched, and before that, we do not believe that OpenAI will be open to the public in a short period of time. However, regardless of when SORA was launched, it is a stronger proof that AI is gradually integrated into human life.
There is no doubt that the advent of SORA is a major victory in the field of artificial intelligence. Based on the existing knowledge and world patterns of human beings, coupled with the corresponding self-learning ability, this is the only way for the development of artificial intelligence. AI companies can easily use this model to provide super tools for a wide range of industries.
In the past few years, whether it is the metaverse, VR, or artificial intelligence, various concepts have been booming, but there has been no specific product. Today, the sales of Meta helmets reach 10 million; With the concept of artificial intelligence, Nvidia has a market value of up to 1$7 trillion; Apple also launched its own headset, Apple Vision Pro; OpenAI's AI products are also being continuously updated. It can be said that an era centered on virtual and artificial intelligence has arrived irreversibly.