Sam Altman is an American entrepreneur, angel investor, programmer and well-known blogger. As a Stanford dropout, he led Silicon Valley startup incubator Y Combinator to help start more than 2000 tech start-ups, including Dropbox, Airbnb, Stripe and other unicorns, with a total valuation of more than $100 billion. However, just as Y Combinator became hot around the world and was even seen as a standard of success, Altman suddenly resigned as president in March to focus on OpenAI, a non-profit artificial intelligence (AI) research organization. And under the leadership of Altman, OpenAI began to turn in a new direction. In fact, all the "YC godfather" did was to achieve an ambitious goal of introducing general artificial intelligence (AGI) into reality.

The following is the text of the article:

After 30 days of arduous negotiations, Sam Altman (Sam Altman) finally convincedMicrosoft CorpThe investment of $1 billion in OpenAI, its leading AI laboratory, marks a new direction for OpenAI. In March, Altman resigned as president of Y Combinator, a Silicon Valley start-up incubator that helped Altman move quickly into the Silicon Valley elite. Now, the 34-year-old is the CEO of OpenAI, which he worked with electric car makers in 2015.TeslaAI Lab, co-founded by CEO Elon Musk (Elon Musk).

Hoping to simulate the mechanism of the human brain.

Musk left the lab last year to focus on his AI ambitions at Tesla, Inc.. Since then, Altman has focused on reinventing OpenAI, a non-profit organization, as a for-profit company so it can be more aggressive in seeking financing. Now Altman has found big-name investors to help it achieve its exciting and lofty goals. Altman and his team hope to build universal artificial intelligence (AGI), a machine that can simulate the human brain to perform any task.

AGI still exists in science fiction. But in the agreement signed by Microsoft Corp and OpenAI, both sides discussed the possibility of making it a reality in the same factual language and applied it to any other technology they wanted to build, whether it was a cloud computing service or a newRobotArm. "my goal of running OpenAI is to successfully create a wide range of useful AI, and the partnership with Microsoft Corp is the most important milestone to date," Altman said in a recent interview.

In recent years, a small but enthusiastic community of AI researchers has set its sights on AGI, and they are supported by many of the world's richest companies.Alphabet Inc-CL CDeepMind, the parent company's top AI laboratory, says it is pursuing the same goal. Most experts believe that AGI will not come for decades or even centuries. Even Altman admits that OpenAI may never achieve its goal, but the competition continues.

In a joint telephone interview with Altman, Microsoft Corp CEO Satya Satya Nadella later compared AGI to his company's efforts to build quantum computers that would be much faster than today's machines. "whether it's our pursuit of quantum computing or our pursuit of AGI, I think you need these ambitious," Nadella said.PolarisThe project will guide the way. "

But since its inception, OpenAI has made many impressive achievements. It has developed open source software that allows other researchers to train and test their AI systems. It has invented a new software that can be learned from trial and error, rather than from existing data. As one of its most famous achievements, it has created five separate AI that can team up to play the complex video game Dota 2. In fact, these AI performed very well, beating many of the world's top human teams.

What makes OpenAI's technology shine in turret 2 is a mathematical technique called reinforcement learning, which allows machines to understand tasks through extreme trials and errors. By playing games over and over again, as well as automated software fragments (called agents), it can learn which strategies can help achieve success. These "agents" learned a lot of skills in a few months and accumulated the equivalent of 45, 000 years of gaming experience.

OpenAI also taught a robot to manipulate objects with unprecedented dexterity. Recently, it has created an algorithm that uses one or two sentences of human-written text, then automatically generates a novel and maintains the same style of writing.

It costs a lot of money to develop new technologies.

However, these breakthroughs are hard to come by. Some of them use a lot of computing power, which means it costs a lot of money. OpenAI has said it suspects that the way to get AGI is to increase the amount of data used in the experiment and need more servers to train the algorithm. To this end, OpenAI spent millions of dollars renting tens of thousands of computer chips, which are located in Alphabet Inc-CL C andAmazonIn the cloud computing services operated by companies. In addition, the "human brain" behind AI is also expensive, and the salaries of many top machine learning experts are eye-popping.

Although OpenAI received a donation of $1 billion, it felt it needed more money. Compared with corporate AI labs, this non-profit organization is at a significant disadvantage. Many enterprise AI labs are part of companies with large-scale cloud computing businesses, so you don't have to pay high fees for server time. In addition, as the largest initial donor, Musk resigned from the OpenAI board in early 2018, reducing investment in OpenAI. OpenAI said Tesla, Inc. was increasingly interested in commercializing AI applications, which created a conflict of interest for Musk.

In order to compete with large companies and startups with huge venture capital reserves, OpenAI's board made an important decision: if it can't beat these companies, join them. In March, Altman announced that OpenAI would create an independent profit-making company called OpenAI LP to seek outside capital. Then this week, it received a huge investment from Microsoft Corp. The terms of the deal have not yet been disclosed, and the co-operation has baffled many observers. "Microsoft Corp already owns one of the most respected private research institutions in the world, the Microsoft Corp Research Institute," Reicht said. "

In response to an email question, Greg Brockman Brockmann, former chief technology officer of payments start-up Stripe and now chairman of OpenAI, said Microsoft Corp had "a large but not a majority" stake in the for-profit entity. Microsoft Corp also has the right to a board seat in the non-profit body, but it has not yet exercised that right, he added.

Microsoft Corp said that his Azure cloud computing infrastructure will be the preferred platform for OpenAI research, so some of his $1 billion cash investment may return to Microsoft Corp in the form of cloud computing. But the two companies also said they would work together to develop a new AI supercomputer.

Perhaps most importantly, Microsoft Corp says it will be the preferred partner for OpenAI, commercialising any technology developed by OpenAI other than AI. Brockmann said the arrangement did not contradict OpenAI's anti-founding philosophy.

"OpenAI has the right to choose, but it is under no obligation to license any particular technology," he said. The purpose of authorizing some of our technologies is to fund our efforts to build AGI so that we can accomplish our mission. But if licensing certain AGI technologies violates the mission, we will not do so. "

Altman and others believe that developing AGI is a huge project that will never be cheap and may require billions of dollars. "in the next few years, we may need to invest in large-scale cloud computing to attract and retain talent and build AI supercomputers," he said. "

There is a long way to go to build AGI

Eventually, Altman and his colleagues believe they can build AGI in a similar way to turret 2. If they can collect enough data to describe the various tasks that humans deal with every day, and have enough computing power to analyze all this data, they believe they can enhance human intelligence.

Altman portrayed the deal with Microsoft Corp as a big step in that direction. While Microsoft Corp invests in OpenAI, the technology giant will also work on building new computing systems to help laboratories analyze more and more information. "this is about building a really tight feedback cycle between our ambitious AGI pursuit and our core business, which is making world computers," Nadella said.

This work may include computer chips specially designed to train AI systems. Like Alphabet Inc-CL C, Amazon.Com Inc and dozens of start-ups around the world, Microsoft Corp is already exploring the new chip.

Since AGI does not yet exist, OpenAI is starting with a smaller project and has recently developed a system that tries to understand natural languages. This technology can meet everything from digital assistants such as Alexa and Google Home to automatic document analysis software within law firms, hospitals and other enterprises.

The deal is also a way for the two companies to promote themselves. OpenAI needs computing power to achieve its ambitions, but it must also attract the world's leading researchers, which is difficult to do in today's job market. Microsoft Corp is competing with Alphabet Inc-CL C and Amazon.Com Inc in the field of cloud computing. In the field of cloud computing, human AI capability is becoming more and more important.

The question is how we should take AGI's ideas seriously. Like everyone else in the technology industry, Altman often says that the future of AGI seems unstoppable. "I think AGI will be the most important technological development in human history," he said. Altman hinted at the concern of Musk and others that AGI might be out of our control. In response, he said: "finding ways to control will be one of the most important social challenges we face."

But games like Tower 2 are a far cry from the complexity of the real world. In recent years AI has made remarkable progress thanks to many technologies developed by companies such as DeepMind and OpenAI. Some systems can recognize images and spoken words and translate them accurately between languages. But this does not mean that AGI is about to become a reality, or even that it is possible. But Oren Etzioni, chief executive of the Seattle-based Allen AI Institute, said: "We are closer to AGI than ever before."

Recently, Alphabet Inc-CL C researcher Geoffrey Hinton (Geoffrey Hinton) won the Turing Award for his contribution to AI research over the past few years. Recently, when someone asked him a question about the AGI competition, he said, "this is too general a question. I prefer to focus on things where you can come up with a solution." Another question about AGI, he added, is: why do we need it?Whether the promise can be fulfilled is questioned.

Founded less than four years ago, OpenAI took a brave stand against the idea that powerful new technologies must be controlled by large enterprises. At the time, the nonprofit's goal was to "develop artificial intelligence in a way that is most likely to benefit all mankind, unfettered by the need to generate economic returns."

Earlier this week, however, OpenAI, which looked very different from its startup, accepted a $1 billion investment from Microsoft Corp, one of the world's most valuable technology companies. The deal gives OpenAI a huge source of funding to compete with well-funded competitors (such as Alphabet Inc-CL C,Facebook Inc, Tencent andBaiduEtc.) to compete.

Under the leadership of Altman, OpenAI has now established a profit-oriented department. As for the opposition to the founding idea of big companies controlling advanced technology, the situation seems a bit delicate. So how to benefit mankind? This is still the mission of OpenAI, but Microsoft Corp needs to make some money in the process now.

As a non-profit, OpenAI has received $1 billion in donations from a number of wealthy supporters, the most prominent being Elon Musk, chief executive of Tesla, Inc.. According to a biography of Musk written by Bloomberg writer Ashlee Vance, Musk is worried that AI could get out of control, including "an army of AI-enhanced droids that can destroy humans."

Musk is also concerned that a small number of companies own intellectual property rights to General artificial Intelligence, or AGI for short. AGI means technology that performs well in a variety of tasks like humans, rather than today's narrow AI that can only do tedious tasks, such as playing chess or recognizing faces.

OpenAI echoed Musk's worldview in his first blog, saying: "as a non-profit, our goal is to create value for everyone, not shareholders." We believe that AI should be an extension of personal will and be deployed as widely and evenly as possible in a spirit of freedom. "

While OpenAI's philosophy feels good, some are sceptical about whether the organization is the right vehicle to democratize the benefits of AI. One of his critics is Ben Recht, a professor at the University of California, Berkeley, who specializes in machine learning algorithms. He is worried about the key roles played by Musk and Altman. "anyone who believes that Musk and Altman benefit mankind in non-profit organizations is deceiving themselves," he said. "this is a weird Silicon Valley vanity project."

Joining OpenAI is not just for money.

So why did Altman resign as president of Y Combinator and join OpenAI as chief executive?

AGI is an interesting bet because it doesn't exist yet, and there is a lot of debate in the tech industry about when AGI will become an implementation, with some thinking it will happen soon, others thinking it may never happen.

According to Altman, "We are working hard to build a secure AGI." So I believe that in the coming decades, someone or some people will build software systems that are smarter and more capable than human beings in all respects. At first, it may be only slightly stronger than humans, but over time, it may be 1 million or even 1 billion times more powerful than humans. So while we are trying to figure out how to implement AGI technically, we also want to ensure its security and fairness and share its benefits. "

Asked how OpenAI plans to make money, Altman replied, "the honest answer is that we don't know." We have never made any income, and we have no plans to generate revenue at present. We don't know how revenue will be generated in the future. " However, Altman has made a soft commitment to investors: "once we have established the AGI system, basically we will let it find a way to bring you a return on investment."

Altman wants OpenAI to set an example for other companies to replicate the so-called "profit capping" (capped profit) business model. OpenAI said its first profit supporters had a profit ceiling of 100 times the initial investment and that any excess profits would flow back to the non-profit organisation. The company said profit caps for subsequent investors would be different.

It is worth noting that 100 times return is a very high threshold, and most investors who invest in ordinary for-profit companies rarely get such a high return. Institutional investor Sequoia Capital (Sequoia Capital) is said to have invested $60 million in WhatsApp, and when the latter was acquired by Facebook Inc for $22 billion, Sequoia Capital made a 50-fold return, which is already a staggering figure.

OpenAI said Microsoft Corp did not participate in the first round of financing, which included Khosla Ventures and Kesla Ventures.LinkedInCo-founder Reed Hoffman's charitable foundation. Hoffman is also a co-founder of OpenAI, but the company declined to disclose Microsoft Corp's upper limit. OpenAI said that in accordance with the requirements of the company's charter, most of the value will eventually belong to all mankind. But at the same time, OpenAI could make Microsoft Corp, already one of the richest companies in the world, richer.

Therefore, the first thing to do is to build an AGI, and the second is to use it to make money. Mr Altman insists: "I really believe that the work we do at OpenAI will far exceed not only what I do at YC, but also what anyone in the technology industry can do." (Golden Deer)