Top of the big era

Zhou Buqi said angrily: "Stop trying to seduce me. I've given you a lot of trouble this morning, and you're still here?"

Shen Xiangyang laughed, then his face straightened, and he said solemnly: "It's the one you mentioned, a general-purpose generative large model, a comprehensive upgrade of Rokid Assistant."

"ah?"

Zhou Buqi suddenly became serious.

This is indeed a big project.

It is even a sign of the beginning of the AI ​​era in the next few years.

Zhou Buqi was a little surprised, "Is the technology mature now?"

"Mature?" Shen Xiangyang was speechless. "It's still early to wait for maturity! However, now the basic direction of the new era of AI has basically been established, deep learning of neural networks. This is already a general consensus in the industry. With this consensus, there is further exploration. In fact, all inventions and innovations are about constant trial and error.”

Zhou Buqi nodded, "It's true!"

Shen Xiangyang said: "The general-purpose generative large model is indeed a huge project and a huge challenge. However, if you don't try, it will always be a challenge. For all technological progress, the first step is It’s all about starting the project. Only when the project is started will talents, funds and even ideas follow.”

"Should it start now?"

"Get ready to start. After the IPO, we will start when we have money."

"Well, that's right."

Zhou Buqi was really afraid that he would be too hasty and rush to launch such a large AI project, which would be too hasty. Such a large risk may not be recognized by the capital market, and it may affect the issuance of stocks.

Now for Ziweixing International, the most important thing is the IPO.

Everything else has to give way.

Shen Xiangyang asked with a smile: "So, is my plan okay? Let's prepare it now and wait until half a year before it is officially launched."

"Yes!" Zhou Buqi agreed with this idea. "The general generative large model is my most important plan in the field of artificial intelligence in the past few years. The ultimate goal of developing Rokid-go, the chess-playing software, is also technical. Accumulate to do this. One thing is that the cooperation between Yahoo and Microsoft will expire next year. By then, there will inevitably be some disputes about whether Yahoo will continue to outsource search to Microsoft or whether Yahoo will do it itself.”

Shen Xiangyang immediately said: "Of course we have to do it ourselves! Search is the core business of the Internet, how can we outsource it? We have neither technology nor talent, so it is a shame to outsource the search business."

Zhou Buqi said: "Even if you do it yourself, you won't be able to make anything famous. With Google as a big mountain, it is impossible to get around it. This is different from trying to compete with Baidu by doing micro-dots in China. Google is almost It is impeccable. On the contrary, if it is outsourced to Microsoft, Yahoo can still get more than 100 million US dollars in revenue from Microsoft every year.”

Shen Xiangyang didn't believe that he would outsource such an important business.

With the character of this big boss, if he doesn’t do something, he must do it himself.

Outsourcing to Microsoft just for the advertising share of more than 100 million a year is not shameful enough.

Shen Xiangyang smiled and said, "Then you still have to do it yourself."

Zhou Buqi nodded and sighed, "You have to do it by yourself, but it is impossible to make any improvement in the short term. Not to mention comparing with Google, even Microsoft's Bing cannot compare. Got it.”

Shen Xiangyang said: "I have calculated that in the search market, as long as we have a 2% share, it is enough to maintain the daily expenses of a 150-person search team. Advertising fees here in the United States are very high, and search advertising has always been very high. popular."

"However, we can't just watch Yahoo search remain on the edge of the industry!"

"Um?"

"If you want to compete with Google in the search field, you must use some new methods! You need some hard-core technology, and general generative AI large models are opportunities. Ziweixing International must get ahead of Google and take this big Let’s get the project started!”

Zhou Buqi put forward his request.

Shen Xiangyang feels the pressure, "This is too difficult. Judging from the number of artificial intelligence scientists, Google has more than us."

Zhou Buqi said in a deep voice: "That must be done."

Shen Xiangyang said with a smile: "But now there is good news."

Zhou Buqi asked: "What's the good news?"

Shen Xiangyang said: "In the artificial intelligence industry, general-purpose generative large models are generally not recognized. It would be easier to specialize in products such as chess-playing software, AI image editing, and AI voice recognition. But if we make a full-scale product Large-scale information models in the field require the training of ultra-large-scale data and a large-scale computing platform. Everyone believes that this is not a goal that can be achieved in the short term.”

Zhou Buqi smiled and said: "That's good. Everyone in the industry thinks it's impossible. This is our opportunity. In a market economy, the more difficult things are, the easier it is to do, and the easier things are, the harder it is to do."

"Uh..." Shen Xiangyang coughed, "I'm talking about the entire industry. The entire industry does not recognize this direction."

"Um?"

Zhou Buqi raised his eyebrows.

Shen Xiangyang said helplessly: "Not only are Google's AI scientists not approving of this direction, our Ziweixing scientists are also not optimistic about this direction. After I proposed this plan, everyone opposed it."

"No one approves?"

Zhou Buqi frowned slightly.

Shen Xiangyang said: "This is what I told you about scientists evading responsibility. This is a super large project, and investing billions of dollars may not be enough. The investment is so large, and the foreseeable hope is slim. Once it fails, no one can bear such a risk."

Zhou Buqi was a little confused, "What kind of reason is this? How can innovation be risk-free?"

Shen Xiangyang said: "Innovation does have risks, but it is related to expected achievements and personal achievements."

"Personal achievements?"

"If you participate in such a large project, the company will invest billions of dollars, tens of billions of dollars, and scientists will have to spend many years to invest in it. The prime years of scientific talents are actually only seven or eight years. If they are all spent on such a large project...and this large project is full of variables, the probability of success is extremely low, and there is also the huge pressure of billions of dollars in investment losses. It is difficult for anyone to afford it. Even for personal reasons, no one wants to achieve nothing in their profession. For individuals, the meaning of innovation is to leave a mark, but this kind of large project can easily drown a person's mark."

"Really."

Zhou Buqi felt the same way.

I have great respect for the scientists who have participated in some super projects over the past hundred years, such as the atomic bomb project and the aerospace project. Not to mention that if they fail, all their life's learning may be wasted. Even if they succeed, many people will remain unknown for their whole lives.

Their own achievements are drowned in super projects.

Innovation in the IT industry is innovation in small projects. It is easier to achieve results and highlight individuals by making a small product or developing a small patent. However, large projects can easily bury many excellent people. Not to mention failure, even if they succeed, it is difficult to stand out.

Shen Xiangyang said: "Now we need to find someone with charisma, appeal and the ability to take on huge responsibilities to lead this project."

"You can't do it?"

"I can't do it!"