Tech

Dawn of the Vincent Video Era: AI Revolution or Anxious Evolution?

On February 16, OpenAI in the United States released the Vincent video model Sora. According to the introduction on the OpenAI official website, Sora can generate smooth virtual videos based on text input by humans. This advancement in video technology has once again stimulated discussions from all walks of life on the impact that artificial intelligence (AI) will have on human society. While public opinion is hotly discussing the bright prospects of the AI ​​era, it is also worried about whether human society has enough ability and resilience to adapt to technological changes. This feeling of anticipation and anxiety has long been accompanied by people’s discussions on the development of AI technology and its impact.

Anticipation and anxiety at the moment of Vincent’s video “GPT3”

In 2023, the generative video models developed by the industry mainly include Runway’s Gen-2, Stability’s Stable Video Diffusion, and Google’s Video Poet. Most of these models only support the generation of videos with a duration of 2 to 20 seconds, while Sora breaks through the bottleneck of generating “60-second ultra-long videos” with current large visual models. From this perspective, Vincent Video seems to have entered the “GPT3 era” relatively quickly.

AI technology will accelerate iteration and improve social efficiency. Zhou Hongyi, founder and chairman of 360 Group, believes that the emergence of Sora means that the time for the emergence of general artificial intelligence will be shortened from ten years to one year. Ben Gozer, an American scientist known as the “Father of General Artificial Intelligence,” made a more optimistic assessment that AI is about to usher in an exponential “intelligence explosion” and reach the level of human intelligence in the next three to eight years. The enabling effect of current Vincent Video technology may be more reflected in improving work efficiency. Algorithms can be used to autonomously learn to quickly synthesize scenes, characters and actions of film and television works, thereby reducing the cost of automated editing, special effects rendering and other tasks. In particular, Vincent Video technology allows creators who lack experience and resources to produce stunning, high-quality videos. Judging from this path of technological empowerment, an era in which “everyone is a film and television creator” is coming.

At a time when disruptive technological concepts are constantly emerging, the uncertainty generated by technological changes has also given rise to a certain kind of anxiety. The rapid development of large AI models, from realizing smooth human-machine dialogue to generating videos, has once again triggered concerns about the impact of emerging technologies on some traditional industries, increasing unemployment and employment imbalances, and exacerbating the proliferation of false information and other challenges to social stability. From the perspective of inter-country relations, the “strongest goes first” development characteristic of AI will also intensify international competition pressure. At the 2024 World Government Summit, NVIDIA founder and CEO Jensen Huang proposed the concept of “Sovereign AI” (Sovereign AI). He said that each country needs to have its own AI infrastructure in order to protect its own cultural and economic security.

Although every country has the need to safeguard its own “sovereign artificial intelligence” interests, large and small countries face varying degrees of vulnerability due to imbalances in power. Large countries have stronger technological reserves and risk resistance capabilities, and can occupy a core position in international technology exchange networks. In contrast, small countries may become increasingly dependent on the technology supply of large countries due to lack of sufficient resources, thus being marginalized in the development of the global AI ecosystem. Technologically weak countries may passively widen the power gap with technologically powerful countries, while technologically powerful countries may pursue excessive securitization policies in order to compete for international core status. This will further increase the intensity of digital geopolitical competition and shrink the already limited space for easing relations between major powers.

Machines replacing humans?

Since AlphaGo defeated Korean Go master Lee Sedol in 2016, while the development of AI technology has entered the fast lane, it has also caused public opinion circles to worry about whether humans can adapt to rapid changes. Some people have even proposed the sci-fi concept of “AI enslaving humans”. , replace human beings” radical assertion. These remarks reflect people’s insecurity in the face of the changes of the times, and everyone is increasingly eager to find paths and solutions to deal with many challenges in a future full of uncertainty.

The current Vincent video technology is still far behind human intelligence. Some technical experts believe that AI technology does not have the human ability to understand and deal with complex problems. Turing Award winner and Meta chief AI scientist Yang Likun said that the ability of an artificial intelligence model to generate realistic videos does not equate to its ability to understand the world like humans. He believes that current AI technology is still difficult to generate videos based on understanding the cause-and-effect relationships in the physical world. Liu Wei, director of the Human-Computer Interaction and Cognitive Engineering Laboratory of Beijing University of Posts and Telecommunications, made a similar judgment. In his view, Sora’s core model is very similar to the GPT large language model. They are both software systems based on conversion architecture, and there are technical obstacles that prevent them from deeply understanding humanities, arts, and moral ethics. From this point of view, these two large AI models have strong learning capabilities, but they are far from the level of processing complex things.

People should look rationally at the impact of AI technology applications on social development. Although there are optimistic expectations in the scientific and industrial circles that AI that can reach the level of human intelligence will eventually enter the public eye, current artificial intelligence technology has not yet shown the effect of triggering disruptive changes in the social structure. From ChatGPT to Sora, the main enabling role played by AI is to improve users’ work efficiency, and it has not yet reached the level of change where “a large number of machines replace human employment”.

Of course, it would be irresponsible to ignore the risks that may arise from the disordered development of AI technology. American computer scientist and futurist Raymond Kurzweil declared that by 2029, AI’s intelligence will reach human levels. “Some people think that my judgment is conservative.” Elon Musk reposted this comment on social media and said that “artificial intelligence may be smarter than any human being next year, and by 2029, it may be smarter than all humans combined.” If AI continues to improve its capabilities through the current “violent computing” method of continuously adding computing power, and eventually achieves “intelligence spillover” that surpasses humans, then the potential risk of allowing the unlimited development of AI technology is extremely huge, and this includes machines. It may create a sense of self-security and protection, and then take actions to get rid of human control, thus controlling, enslaving or even annihilating humans.

From the perspective of regulating the potential risks of artificial intelligence, countries should promote cooperation in AI global governance while safeguarding the interests of “sovereign artificial intelligence”. Related governance in the field of AI is related to the destiny of mankind and is a serious issue facing global governance. The international community needs to pay attention to the vulnerability of developing countries in the wave of emerging technologies. Technologically powerful countries should not blindly adopt the method of “decoupling and breaking links” just to consider their own national interests. AI is a dual-use technology for both military and civilian use. As far as its civilian application is concerned, it is in the common interest of all countries to guide the healthy development of the global AI industry and realize “AI for good”, thereby promoting AI technology to benefit all mankind. In terms of more sensitive military security, the militarized application of AI has increasingly attracted the attention of international public opinion. According to international media reports, AI technology has been used in the new round of Palestinian-Israeli conflicts, causing civilian casualties. From the perspective of constructing global security governance rules for artificial intelligence, the international community may need to explore the establishment of some kind of common proposition of “humane AI” to avoid humanitarian disasters caused by algorithms determining life and death.

In short, to promote the realization of the goal of “AI for mankind”, we need to be guided by the concept of a community with a shared future for mankind, establish the awareness and actions of “global artificial intelligence”, and promote the coordinated advancement of technological progress and empowerment, risk prevention and the formulation of international norms. , complement each other.

error: Content is protected !!