Just received 2 billion. Deduction of the endgame of the CEO of a unicorn valued at tens of billions: Humans only have two things left.

Embodied AI has become the forefront of global technological competition. Amid this wave, today marks a significant industry milestone: Galaxea, led by Gao Jiyang, officially announced the completion of a nearly 2 billion yuan Series B+ financing round, with its valuation exceeding 20 billion yuan at once, breaking the valuation record of unlisted embodied AI companies in China.

In the current era of capital flow and technological leapfrog, in the 9th episode of "Shanyou Exploration Stream", Professor Li Shanyou invited the founder and CEO of Galaxea, who is now in the spotlight of the entire industry, for an in - depth dialogue. This podcast is very interesting. It is a "cognitive conversation" full of sparks between two minds.

Gao Jiyang is a typical entrepreneur and scientist. He is rational, pragmatic, and has a strong "productivity" mindset and a sense of mission in the era. In the program, he delved into the industry, hoping to build a future with "rationality".

Professor Li Shanyou used Socratic questioning, constantly challenging and prying Gao Jiyang's rational framework with more ultimate questions (the uniqueness of humanity, the boundaries of rationality, and the "abstract thinking" of the founder), forcing him to step out of the "CEO" role and examine the philosophical foundation of himself as a "creator". In Professor Li Shanyou's view, "the founder is not a well - rounded warrior but a Prometheus who steals fire."

The interaction between the "pragmatic scientist" and the "first - principle - based questioning" generates a great deal of cognitive tension.

This podcast is like a thinking documentary, recording how a doer at the forefront of the AI wave examines the abyss and the stars that the wave may ultimately head towards. It is worth listening to for those who are concerned about technology, business, philosophy, and the future of humanity.

The highlights of the dialogue are organized as follows:

01

Will there be a "Zhang Yiming" in the AI era?

Professor: Your experience is very clear. You participated in the Olympiad, studied at Tsinghua University, and went abroad for three and a half years to pursue a doctorate. Is there anything in your academic experience that others don't know?

Gao Jiyang: My intelligence is not the kind of Tsinghua top - student type. It was only when I graduated from university, started my doctorate, and entered society that I felt like a fish in water. Before that, I was on a very narrow track, only competing with others in one dimension. Although the results at each stage seemed good, the process was not as smooth as people thought.

Professor: What are the situations of young people in AI and embodied AI laboratories in China and the United States now?

Gao Jiyang: In the United States in the past few years, there was a characteristic that the Silicon Valley had a rather strict hierarchical structure. Coupled with the English - speaking environment, Indians were very prominent in the wave of the Internet. As a result, many talented young Chinese people had difficulty getting promoted and their work status was more like that of pure engineers.

But things have changed in the past two years. The strict hierarchical team structure doesn't work anymore. It still depends on those who really understand large models to lead a capable team to train the basic models. In OpenAI, Meta, and even Google's Gemini team, the proportion of Chinese people is very high.

In China, whether we are doing embodied AI or large models, it is still those born around 1990 and 1995 who are really leading the teams at the front line. I think this is the task entrusted to us by the era, and I just happened to catch up with it.

Professor: In the AI era, the post - 90s generation has emerged. Do you think there will be a new Zhang Yiming?

Gao Jiyang: There will definitely be, but the form in which it appears will probably be beyond everyone's imagination.

Many investors also ask me what will happen if large - scale enterprises enter our field. At this time, I usually make an analysis: Under what circumstances will large - scale enterprises suppress start - up companies? Look at it from three aspects: one is organization, the second is capital and resources, and the third is business synergy. In many cases, the absolute advantage of large - scale enterprises lies in the third point.

Looking at embodied AI again: On the supply side, almost all the supplies are new. When it comes to making robots, joints, and overall machine designs, these things didn't exist before. In terms of data, there is more operational and motion - related data. Large - scale enterprises have no such data at all. As long as you are fast, you can complete the accumulation first. On the demand side, embodied AI ultimately provides productivity in the physical world. In fact, no one can monopolize the productivity in the physical world. So from both the supply and demand sides, start - up companies in the field of embodied AI do not have a disadvantage in business synergy when competing with large - scale enterprises.

As for organization, in the new era, both humans and AI are equal parts of our organizational system. Facing the transformation of enterprise productivity, small companies may have more advantages.

In terms of capital and talent, I think leading start - up companies do not lack capital in such undertakings. And the best talents still want to make good products. Those who can make good products will naturally attract these people to stay.

Is it possible for super - large enterprises to emerge in the AI era? I think there is a great chance, at least in the field of embodied AI.

Professor: What will be the difference between the "Zhang Yiming" in the AI era and the Zhang Yiming back then?

Gao Jiyang: Sometimes I wonder what the future society of humanity will be like when AI is truly popularized. I think humans will become the last bottleneck in the productivity system and gradually withdraw from this system.

Embodied AI, that is, AI in the physical world, actually has two major values: one is to change, which will completely restructure the productivity system on the earth; the other is to explore, exploring territories beyond the earth.

The real vast expanse for the founders and entrepreneurs of this generation of embodied AI may be in dimensions beyond the earth.

So the requirements for the founders of such companies actually have two sides. On the one hand, they can talk about AI and the future with others. On the other hand, they have to be rooted in the traditional Chinese manufacturing supply - chain system. I have to be a well - rounded warrior. This portrait is actually very different from that of traditional Internet founders and also completely different from that of large - model founders.

02

The CEO should have the awareness of a fire - stealer

Professor: What is the vast expanse in your heart?

Gao Jiyang: When I was very young, I watched "Kangxi Dynasty" and saw how the young Kangxi ascended the throne, fought against Oboi, and pacified the Three Feudatories... He was a person who had a thirst for making achievements and also had a sense of mission and sentiment. I was deeply touched at that time and thought that I should also make a career in the future. But what does it mean to make a career? I didn't have a concept at that time. Later, as I grew up, I was greatly influenced by Confucianism. I saw that Confucianism talked about the three immortals: establishing merits, establishing virtues, and establishing words, which gave me a lot of inspiration.

I increasingly feel that doing embodied AI or autonomous driving is the mission entrusted to me by this era. If I had lived 20 years earlier, I might have had to do something else.

Professor: I think you are very interesting. You are both humble and not humble. You have placed yourself in the context of the era and then chosen a point to break through. When you set your ambition very high, you suddenly become humble, but the underlying color is a sense of "being in position". Just like the young Kangxi you mentioned, he was humble to others, but behind that humility was a great sense of being in position. This is what makes you different from many people.

Gao Jiyang: I think I am getting along with the world in a humble way. In addition, I am quite persistent in many choices. Once I make a choice, I will start working on it and won't worry about not being able to succeed. If I can't succeed, it means that my accumulation is not enough and my method is not right, not that I am not good enough.

What I care about is whether the world will change without our company.

Professor: What kind of change do you hope to bring to the world because of you?

Gao Jiyang: This definitely has to be done in several steps. The first step is to have a 24 - hour fully autonomous and closed - loop robot system in the scenarios we can see, including logistics and intelligent manufacturing. There are also several important nodes in between. One of the important nodes is to be able to deploy a complete robot productivity system in a remote area. This is a medium - term state.

After many steps, we will find that the material wealth on the entire earth has increased greatly, and people no longer have to worry about work. Further on, as Elon Musk said, exploring the universe, exploring space... If because of our efforts, we become the last generation of humans who need to work, that will be my greatest wish.

Professor: It is very important to find your own position in this world. In the business world, how do you position yourself?

Gao Jiyang: Compared with many entrepreneurs who admire Steve Jobs and Elon Musk, I think I may be more practical. I would like to first become Wang Chuanfu of BYD. BYD is very powerful in battery technology and three - electric technology. It is a very solid enterprise. I think I should first be like Wang Chuanfu and solidly build our industry, and then benefit many people through this industry.

A good state for entrepreneurship is that you can bring value to everyone.

03 What exactly is AI - native?

Professor: We are very lucky to witness the changes in the era. It is not only the technology and the market that are changing, but also the cognitive trends. As a third - generation AI entrepreneur, looking back at the first - generation BAT and the second - generation mobile Internet, what do you think are the differences?

Gao Jiyang: I think ByteDance has a very significant difference from others in that it is very data - driven and almost makes all decisions based on data. In the BAT generation, people also used data, but they didn't take it to such an extreme.

What is the essence of data - driven? I think it is rationality in the absolute sense.

I am thinking that for an enterprise to become a productivity system, there must be inputs and outputs. So what kind of productivity is the strongest? Perfect productivity is definitely completely rational. Of course, rationality doesn't mean not making mistakes. It needs a scientific error - testing mechanism to explore the laws of the world and then adjust its decisions based on the collected data.

I think these AI - related enterprises may need to evolve further and use a more absolutely rational productivity from a more fundamental level. This absolutely rational productivity is AI.

Previously, enterprises emphasized values very much. Why? Because people need to be guided by values to enter an orderly working state. Later, we emphasized data - driven more because many times a truth can be argued both ways, but data doesn't lie.

I believe that in the future, when each of our production units is AI, each role will naturally be data - driven. I think this may be some of the changes that will occur among these generations of enterprises that I have observed.

Professor: Assuming that we all believe that AI - Native is a new era, what exactly is AI - Native?

Gao Jiyang: I think the development of AI itself can be divided into several stages: the first stage is to use AI as a tool; the second stage, which is now, is that AI has become a productivity; we haven't reached the third stage yet, when AI can self - iterate.

An AI - Native organization should essentially complete self - iteration through AI. From this perspective, all enterprises today have not reached this level.

Each of us should think about a question. If AI were to take over my position, what value could I provide for it? Instead of the other way around, what value it can provide for me. There is a basic assumption here, which is that we have to admit that AI is an existence far beyond humans in terms of production capacity and thinking ability.

So at a certain stage, we have to accept that humans should cooperate with AI in production.

What else can humans do? I think the value of humans lies in two points: the first is responsibility. For example, when something goes wrong, someone has to take the blame. The second is judgment. All the taste of AI is actually taught by humans, and human taste is obtained from the physical world. We judge what is good. These may be the last two strongholds of humans.

Professor: You just mentioned a very key word, which is the self - evolution of AI. AI - Native means that AI evolves itself, continuously learns, and is always online.

When AI can self - iterate, the AI it iterates out is not the AI trained by humans at all. Humans have many limitations, such as our computing power and our way of thinking in language. But large models are essentially parameters. Maybe in the future, AI will communicate directly in parameters. If it develops in this direction, it will iterate at the fundamental level with its own language, and we won't know what it is doing at all, not even understand its language.

Pushing this to the extreme, if AI really develops to the point where it can self - learn and iterate, what will be the last stronghold of humanity? Is it still responsibility and taste?

Gao Jiyang: This has to go back to a more grand question - what is the meaning of the existence of the world and civilization? When humans no longer contribute to the basic survival and development of this civilization, it actually means that AI can lead humans to more distant unknown realms.

For example, in the unknown realm of mathematics, many conjectures and hypotheses have not been proven. There are also some possibilities that each person still has many cognitive - level problems that have not been thought through. AI can more rationally tell you what the laws and principles are. This is actually the unknown realm of human nature.

Moreover, I always believe that AI is meaningless without humans. AI always serves humans, and civilization is centered around humans. The frontier of civilization may not be explored by humans in the future, but may be explored by AI instead. However, the meaning of the results they explore lies in allowing humans to see the unknown realm.

04

"I can see two things in your mind fighting with each other"

Professor: You have been using a word to describe AI, which is "absolute rationality", saying that it can take us to understand those unknown realms. But Gödel's incompleteness theorem tells us that even if there were a hundred Einsteins at the same time, humans would never be able to understand the truth of the universe with rationality. Rationality is a matrix, it is a cage, and rationality has boundaries.

If human rationality has boundaries, can we transcend Gödel's incompleteness theorem by adding the absolute rationality of AI?

Gao Jiyang: It definitely won't work in the AGI era. It has to self - evolve, and only what it evolves may be possible.

Professor: It may evolve something beyond rationality, and only then is it possible. If we don't transcend rationality, the unknown realm it takes us to will only be a place that we can't reach the end of.

I'll ask a more direct question. There are two most important philosophies in the world. One is ontology, which asks what the world is. The second is epistemology, which asks about my cognition. Do you believe that there is a higher intelligence beyond rationality? Do you believe that there is a greater existence behind this world?

Gao Jiyang: To some extent, I may not believe it so much. For example, I don't quite believe that the world we live in is just a small world in a multiverse or a large space.

I think people still prefer to believe that they are a unique existence in this world, rather than there being a more unique existence. If there is a more advanced existence beyond humans and humans are just the fuse of a higher civilization, and if most people in the world believe this, human civilization will be in jeopardy.

Professor: I can see two things in your mind fighting with each other. The first is humans, and the second is robots.

You think it's great to let AI become the labor force, but there is an implicit premise that it has to "serve humans and humans dominate human interests". But there is indeed a risk here. If one day robots really surpass humans