Greater than Sora, What AI Breakthrough
Recently, there have been new advances in AI, particularly in the AI video generation model Sora. While previous AI-generated videos were limited to 4s, Sora can now generate 60s videos that closely resemble the real world.
Some scientists and AI experts believe AI has mastered the physical model of the world, enabling it to simulate the real world and understand various physical laws like humans.
However, I believe that AI still has a long way to go before it can truly comprehend the physical world.
AI, still in the animal intelligence stage
Human understanding of the physical world relies on two parts of intelligence.
One is animal intelligence, including a memory prediction model that generates memories of the environment and predicts future behavior to interact better with it.
The emergence of Sora indicates that AI has achieved animal-level intelligence. However, it still has a long way before reaching human-level intelligence, which requires rational thinking to comprehend the world.
We communicate using language and reach a consensus on topics such as Newton's laws and Einstein's theory of relativity. Rationality enables us to summarise universal laws.
And SORA, It's evident that we have not yet achieved generalised AI and are still at the stage of monolithic intelligence.
Intelligence can be irrational, and our big models are trained on a lot of data inputs to develop intelligence, but not rationality.
The Sora model is also intelligent but not rational. It's a visual intelligence, while ChatGPT is a linguistic intelligence. There may be more types of intelligence in the future, but none of them will be rational. It's important to note that the original meaning and quotes have been preserved.
However, combining several models doesn't solve the challenge of generalising the different intelligences of language and vision into a universal one.
Turning point of AI
AI has reached a turning point. Big models can still form stronger intelligence through proprietary data training, but they cannot solve all problems.
This means intelligence can be formed in all industries. There are opportunities for AI applications, as well as methods of using AI applications.
1. Use the synergistic capabilities of humans
Nowadays, AI is already at a medium level of proficiency and has the potential to perform a wide range of tasks.
However, It's important to recognize that the strength of humans lies in our ability to work together synergistically. While no individual may possess all the knowledge necessary to land on the moon, collectively we can achieve great things through complex synergy.
2. Learn to command AI to work
Currently, AI has not been socialized, and as a result, no AI society has been formed. However, the future is moving towards socialization.
Similarly, the core of human intelligence doesn't lie in individual super-intelligence, but rather in the collective intelligence of society. Culture serves as a reservoir of intelligence, allowing us to summarize complex ideas into a communicable language and precipitate them into our cultural heritage.
It's through culture that human society gains its strength.
While many people express concern about AI surpassing human beings, It's important to note that humans are also evolving. Unlike AI, the group evolution of human beings is fast, while the group evolution of AI has not yet begun. Only when AI begins to evolve can it be a match for humans.
Therefore, the value of humans lies in their ability to collaborate with machines and drive them to work more efficiently, resulting in a higher machine IQ.
Biggest opportunity lies in the cross-border collision
Complex services can create barriers to business. With the recent changes in AI, It's important to focus on how to commercialise and make these technologies more widely available. The development of any advanced technology has its own set of rules, and each round of technological advancement is a step towards universal technology. It's important to note that the creation of a complex service can result in the formation of a strong barrier due to the development of a new intelligent model. I believe the next advancement is dependent on AI agent.
It's not the end of an era, but rather the start of a new one.
