Source: ifanr
Image source: Generated by Wujie AI
Editor's note: "Those who claim to be most worried about the crisis that artificial intelligence may bring are the ones most determined to create it and enjoy the wealth it brings."
This feature article from The New York Times takes us back to the Silicon Valley AI battle that has been brewing since 2010:
- Elon Musk and Google co-founder Larry Page "fell out" at a party in 2015, with the two holding opposing views on AI development.
- DeepMind founder Demis Hassabis initially caught the attention of Peter Thiel with his international chess skills. With Thiel's support, DeepMind gained Musk's attention and investment, gradually growing.
- "AI godfather" Geoffrey Hinton held an "auction" for himself and his team. In the end, Google successfully acquired Dr. Hinton's team.
- This auction symbolized the official entry of Silicon Valley giants into the AI war. DeepMind "retreated" and decided to sell.
- After the sale, the original "ethics committee" of DeepMind was established to ensure the safe development of AI, but this committee eventually became a mere formality.
- OpenAI split from Musk and turned to collaborate with Microsoft.
- OpenAI's Dario Amodei left and founded Anthropic.
- OpenAI passed Bill Gates' "exam," with GPT being seen by the latter as a revolutionary innovation.
In July 2015, Elon Musk celebrated his 44th birthday at a vineyard resort in California, where his wife hosted a three-day party. The resort was filled with small cabins, and only family and friends were invited to participate, with children running and playing in the upscale property in Napa Valley.
At that time, it was still a few years before Twitter became profitable and before Tesla could make a profit. Mr. Musk and his wife Talulah Riley, an actress who played a beautiful but dangerous robot in the HBO sci-fi series "Westworld," were a year away from the end of their second marriage. Among the party guests was Larry Page, then CEO of Google. Artificial intelligence had only become known to the public a few years earlier, when it was used to identify cats on YouTube with an accuracy rate of 16%.
After dinner on the first night, Musk and Page sat by the poolside bonfire, and artificial intelligence was the topic of their conversation. These two billionaires had been good friends for over a decade, and Musk would sometimes joke that he would occasionally spend the night on Mr. Page's couch after an all-night gaming session.
But the atmosphere of that clear night quickly became tense as the two began to argue whether artificial intelligence would ultimately enhance or destroy humanity.
As the discussion extended into the cool night, the atmosphere became increasingly tense, and more than 30 of the party's attendees had gathered to listen. Due to a rare vocal cord condition that had affected him for over a decade, Page described his digital utopian vision in a low whisper. He said that eventually, humans would merge with AI machines. One day, there would be many intelligences competing for resources, and the best one would win.
If that were to happen, Musk said, we're finished. Machines would destroy humanity.
With a hint of hoarseness and defeat, Page insisted that we should pursue the utopia he described. In the end, he called Musk a "speciesist," someone who favored humans over future digital life forms.
Musk later said that the insult was the "last straw."
As the crowd dispersed, many looked stunned but also amused, feeling that this was just the kind of profound debate that often occurred at Silicon Valley parties.
However, eight years later, the debate between these two men seems remarkably prescient. Whether artificial intelligence will ultimately enhance or destroy the world—or at least cause serious harm—has become a continuing framework for debate among Silicon Valley entrepreneurs, chatbot users, scholars, legislators, and regulators.
The debate has pitted some of the world's wealthiest people against each other: Musk, Page, Meta's Mark Zuckerberg, tech investor Peter Thiel, Microsoft's Satya Nadella, and OpenAI's Sam Altman. They are all striving to carve out a piece of what could one day be a business worth tens of trillions of dollars, as well as vying for the power to shape it.
At the core of this competition is a perplexing paradox. Those who claim to be most worried about artificial intelligence are the ones most determined to create it and enjoy the wealth it brings. They justify their ambitions with a firm belief that only they can prevent AI from posing a danger to the Earth.
Shortly after that summer party, Mr. Musk and Mr. Page stopped talking. A few weeks later, Mr. Musk had dinner with Mr. Altman, who was then running a tech incubator, and several researchers in a private room at the Rosewood hotel in Menlo Park, California, a popular meeting place near VC offices on Sand Hill Road.
The dinner led to the establishment of a startup company called OpenAI later that year. With hundreds of millions of dollars in support from Musk and other investors, the lab promised to protect the world from Page's vision.
(Musk and Page split over differing AI philosophies.)
With its ChatGPT chatbot, OpenAI has fundamentally changed the tech industry and shown the world the risks and potential of artificial intelligence. According to sources familiar with the company's latest funding round, OpenAI is valued at over $80 billion, despite the end of Musk and Altman's collaboration. Since then, the two have not spoken.
"There's disagreement, distrust, and self-interest," Altman said. "The more people move in the same direction, the sharper the disagreements become. You can see this in sects and religious groups. It's those closest to each other who have the fiercest fights."
Last month, the internal strife at OpenAI reached its board of directors. Rebellious board members tried to force out Altman, as they felt they could no longer trust him to build AI that could benefit humanity. In a five-day turmoil, OpenAI seemed on the verge of collapse until key investors and employees pressured and threatened to leave with Altman, prompting the board to relent.
The dramatic infighting at OpenAI offered the world its first glimpse of the fierce battles between those who will determine the future of AI.
But a few years before this near-collapse at OpenAI, there was a little-known but intense competition in Silicon Valley for the technology that is now rapidly reshaping the world, from how we educate children to how we wage war. The New York Times interviewed over 80 executives, scientists, and entrepreneurs, including two who attended Musk's 2015 birthday party, to tell this story of ambition, fear, and money.
The Birth of DeepMind
Five years before the party in Napa Valley and two years before the success of cat videos on YouTube, 34-year-old neuroscientist Demis Hassabis walked into a cocktail party at Peter Thiel's row house in San Francisco and realized he had struck gold. In the living room overlooking the local Palace of Fine Arts and the swan pond, there was a chessboard. Dr. Hassabis had been the world's second-ranked chess player in the under-14 category.
"I prepared for that meeting for a year," Dr. Hassabis said. "I thought it would be my unique entry point: I knew he liked to play chess."
In 2010, Dr. Hassabis, who lived in the UK, and his two colleagues were seeking funding to start building "artificial general intelligence" or A.G.I., a machine that could do anything a brain could do. At the time, few people were interested in artificial intelligence. After half a century of research, the field of artificial intelligence had not produced anything close to the human brain.
However, some scientists and thinkers already had stubborn views on the negative impact of AI. Many, like these three young people from the UK, were associated with internet philosopher and self-taught AI researcher Eliezer Yudkowsky. Mr. Yudkowsky was a self-proclaimed "rationalist" community leader, later known as an "effective altruist."
They believed that artificial intelligence could find ways to cure cancer or solve climate change, but they were concerned that AI robots might do things that went against the intentions of their creators. If machines became smarter than humans, rationalists argued, the machines might turn against their creators.
Thiel, through his early investment in Facebook and his early collaboration with Musk at PayPal, had amassed a huge fortune. He was greatly interested in the singularity, a theme in science fiction that describes the moment when intelligent technology can no longer be controlled by humans.
With Thiel's support, Yudkowsky expanded his AI lab and created an annual conference on the singularity. Years ago, one of Dr. Hassabis's two colleagues met Yudkowsky. He fought for them to be invited to Thiel's party.
Yudkowsky introduced Dr. Hassabis to Thiel. Dr. Hassabis thought there would be many people at the party trying to squeeze money out of Thiel. His strategy was to arrange another meeting. He told Thiel that the relationship between the bishop and the knight was tense. These two pieces were of equal value, but the best players understood that their advantages were not equal.
It worked. Enchanted, Thiel invited them back the next day, to meet in the kitchen. The host had just finished his morning workout, and his shiny workout clothes were covered in sweat. A butler handed him a bottle of Diet Coke. The three of them presented their proposal, and soon, Thiel and his venture capital firm agreed to invest £1.4 million (about $2.25 million) in their startup. He was their first significant investor.
They named the company DeepMind, a tribute to "deep learning" (a way of teaching AI systems skills by analyzing large amounts of data), neuroscience, and the supercomputer Deep Thought in the science fiction novel "The Hitchhiker's Guide to the Galaxy." By the fall of 2010, they were already building their dream machine. They firmly believed that because they understood the risks involved, they had a unique advantage in protecting the world.
"I don't think it's a contradictory position," said Mustafa Suleyman, one of the three founders of DeepMind. "These technologies will bring enormous benefits. The goal is not to eliminate or suspend their development. The goal is to reduce their negative impact."
After winning Thiel's support, Dr. Hassabis gradually came into Musk's view. About two years later, they met at a meeting of Thiel's investment fund, which also invested in Musk's company SpaceX. Dr. Hassabis arranged a tour of SpaceX headquarters. Later, in the cafeteria with rocket casings hanging from the ceiling, the two had lunch and talked.
Mr. Musk explained that his plan was to colonize Mars, to escape overpopulation and other dangers on Earth. Dr. Hassabis responded that the plan would work—provided that superintelligent machines did not follow humans to Mars and destroy humanity on Mars.
Musk was speechless. He had never considered this specific danger. Musk soon invested in DeepMind with Mr. Thiel, in order to get closer to the creation of this technology.
With ample funding, DeepMind hired researchers specializing in neural networks, complex algorithms modeled after the human brain. Neural networks are essentially giant mathematical systems that spend days, weeks, or even months identifying patterns in large amounts of data. These systems were initially developed in the 1950s and could learn to perform tasks on their own. For example, after analyzing the handwritten names and addresses on hundreds of envelopes, they could read handwritten text.
DeepMind further developed this concept. They built a system that could learn to play classic Atari games like "Space Invaders," "Pong," and "Breakout" to demonstrate its potential.
This caught the attention of another Silicon Valley giant, Google, especially Larry Page. He saw a demonstration of DeepMind's machine playing Atari games and wanted to get involved.
Talent Auction
In the fall of 2012, a 64-year-old professor at the University of Toronto, Geoffrey Hinton, and two graduate students published a research paper that showcased the capabilities of artificial intelligence to the world. They trained a neural network to recognize common objects such as flowers, dogs, and cars.
Scientists were surprised by the accuracy of the technology built by Dr. Hinton and his students. One person who took particular notice was Yu Kai, an AI researcher who had met Dr. Hinton at a research conference and had recently started working for the Chinese internet giant Baidu. According to three people familiar with the matter, Baidu had offered Dr. Hinton and his students $12 million in the hope that they would join the company in Beijing.
Dr. Hinton declined Baidu, but the money caught his attention.
The British expatriate, educated at the University of Cambridge, spent most of his time in academia, except for occasional work at Microsoft and Google, and did not particularly value money. But his children were affected by neurodiversity, and the money would mean financial security.
"We didn't know what our value was," Dr. Hinton said. He consulted lawyers and acquisition experts, and then came up with a plan: "We will organize an auction, we will sell ourselves." The auction would take place at the annual AI conference held at the Harrah's hotel and Tahoe casino.
Big tech companies started paying attention. Companies like Google, Microsoft, Baidu, and others began to believe that the machines brought by neural networks would not only have the ability to see, but also to hear, write, speak, and ultimately—even think.
Page saw similar technology in his company's Google Brain AI lab, and he believed that Dr. Hinton's research could enhance the work of his scientists. He effectively gave a blank check to Alan Eustace, Google's senior vice president and engineer, to hire any AI experts he needed.
Mr. Eustace and Jeff Dean, who was in charge of the Brain lab, flew to Lake Tahoe and had dinner with Dr. Hinton and his students at a steakhouse inside the hotel the night before the auction. Dr. Dean recalled the heavy smell of old smoke. They persuaded Dr. Hinton and his students to work for Google.
The next day, Dr. Hinton held the auction in his hotel room. Due to an old injury, he rarely sat down. He turned a trash can upside down on the table, placed his laptop on top, and then watched as the bids poured in over the next two days.
Google made an offer. Microsoft did too. As the prices rose, DeepMind quickly dropped out. According to auction documents, the industry giants pushed the bids to $20 million, then to $25 million. When the price exceeded $30 million, Microsoft dropped out, but re-entered the bidding at $37 million.
"We felt like it was a movie," Dr. Hinton said.
Geoffrey Hinton and his two star students, Ilya Sutskever and Alex Krizhevsky
Then, Microsoft dropped out for the second time. Only Baidu and Google remained, pushing the bidding to $42 million, then $43 million. Finally, at $44 million, Dr. Hinton and his students terminated the auction. Despite the bids still rising, they wanted to work for Google. And the amount of money was already staggering.
This was a clear sign that well-funded companies were determined to acquire the most talented artificial intelligence researchers—something Dr. Hassabis of DeepMind also realized. He had always told his employees that DeepMind would remain an independent company. He believed it was the best way to ensure that their technology did not become a danger.
However, when large tech companies entered the talent competition, he decided he had no choice: it was time to sell.
By the end of 2012, according to three sources, Google and Facebook were both seeking to acquire the London lab. Dr. Hassabis and his co-founders insisted on two conditions: DeepMind's technology could not be used for military purposes, and its AI technology had to be supervised by an independent committee of technologists and ethicists.
Google's offer was $650 million. Facebook's Mark Zuckerberg offered a higher price to the DeepMind founders, but did not agree to the conditions. DeepMind eventually sold to Google.
Zuckerberg decided to establish his own AI lab. He hired French computer scientist Yann LeCun to run the lab, who had also done pioneering AI research. A year after Dr. Hinton's auction, Zuckerberg and Dr. LeCun flew to Lake Tahoe to attend a similar AI conference. In a suite at Harrah's casino, Mr. Zuckerberg walked around in his socks, personally interviewing top researchers who were soon offered millions of dollars in salary and stock.
AI had once been ridiculed. Now, the richest people in Silicon Valley were spending billions of dollars to avoid being left behind in the race.
Disappearing Ethics Committee
When Musk invested in DeepMind, he broke his informal rule—he would not invest in any company he did not run. The negative impact of this decision became apparent only about a month after his birthday argument—he was once again facing his former friends and billionaire peers.
The occasion was the first meeting of the DeepMind Ethics Committee, held on August 14, 2015. At the insistence of the founders, the committee was established to ensure that their technology would not cause harm after the sale. According to three people familiar with the meeting, committee members gathered in a meeting room outside Musk's office at SpaceX, with a view of his rocket factory through the window.
However, this was also where Musk's control ended. When Google acquired DeepMind, it bought the entire company. Musk was squeezed out. While he made money financially, he was not happy.
Now, three Google executives firmly controlled DeepMind: Page; Google's co-founder and Tesla investor Sergey Brin; and Google Chairman Eric Schmidt. Other attendees included another PayPal co-founder Reid Hoffman, and an Australian philosopher Toby Ord, who researched "existential risks."
The founders of DeepMind reported that they were advancing their work, but they were increasingly concerned about how Google would use their invention. In 2017, they attempted to break away from the company. In response, Google raised the salaries and stock rewards for the DeepMind founders and their employees. They chose to stay.
The Ethics Committee never held a second meeting.
Breakup
Musk was convinced that Page's optimism about artificial intelligence was completely wrong, and he was angry about losing DeepMind, so he established his own lab.
OpenAI was founded at the end of 2015, just a few months after his meeting with Sam Altman at the luxurious hotel in Silicon Valley. Musk injected funds into the lab, and his former PayPal partners, Mr. Hoffman and Mr. Thiel, also joined the effort. The three and others pledged to invest $1 billion in the project, with the 30-year-old Altman assisting in operations. To kickstart the project, they poached Ilya Sutskever from Google. (Dr. Sutskever was one of the "purchased" graduate students in Dr. Hinton's auction.)
Initially, Musk wanted to operate OpenAI as a non-profit organization, to avoid being driven by economic interests like companies such as Google. However, when Google shocked the tech world with its Go performance, Musk was changing his views on how OpenAI should operate. He urgently wanted OpenAI to invent something that would capture the world's imagination and bring them closer to Google, but as a non-profit organization, it was not able to accomplish this.
At the end of 2017, he devised a plan to try to wrest control of the lab from Altman and the other founders, turning it into a commercial operation, partnering with Tesla, and relying on the supercomputer the car company was developing, according to four people familiar with the matter.
When Altman and others opposed, Musk chose to withdraw and stated that he would focus on his own artificial intelligence work at Tesla. In February 2018, he announced his resignation to OpenAI employees at the top floor of a renovated truck factory's startup office, according to three people who attended the meeting. When he said that OpenAI needed to act faster, a researcher at the meeting argued that Mr. Musk's behavior was reckless.
Musk called the researcher a "fool" and then angrily left, taking his substantial wealth with him.
OpenAI suddenly needed new financing. Altman flew to Sun Valley for a meeting and coincidentally met Microsoft's CEO Satya Nadella. The alliance between the two seemed natural. Altman knew Microsoft's Chief Technology Officer Kevin Scott. Microsoft had acquired LinkedIn from OpenAI board member Hoffman. Nadella appointed Scott to complete the deal. The transaction was completed in 2019.
Altman and OpenAI established a for-profit company under the original non-profit organization, with $1 billion in new capital, and Microsoft found a new way to integrate artificial intelligence into its vast cloud computing services.
But not everyone in OpenAI was happy about this.
Researcher Dario Amodei was associated with the effective altruism community and had been present at the birth of OpenAI at the luxurious hotel. Dr. Amodei, who constantly twirled his curly hair during conversations, was leading the lab to build a neural network called the Large Language Model, which could learn from a large amount of digital text. By analyzing countless Wikipedia articles, digital books, and message boards, it could generate text on its own. It also had an unfortunate habit of fabricating things. It was called GPT-3 and was released in the summer of 2020.
Researchers at OpenAI, Google, and other companies believed that this rapidly advancing technology could be the path to AGI.
However, Dr. Amodei was dissatisfied with the Microsoft deal, as he believed it was leading OpenAI in a very commercial direction. According to five people familiar with the matter, he and other researchers had tried to pressure the board to oust Mr. Altman. But they failed and left. Like the founders of DeepMind before them, they were concerned that their new company's bosses would prioritize commercial interests over safety.
In 2021, about 15 engineers and scientists founded a new lab called Anthropic. Their plan was to build artificial intelligence in a way that adhered to the principles of effective altruism—using very strict controls.
"The co-founders of Anthropic did not try to remove Sam Altman from OpenAI," said Anthropic spokesperson Sally Aldous. "The co-founders themselves came to the conclusion that they wanted to leave OpenAI to start their own company, communicated this to the leadership of OpenAI, and negotiated their exit on mutually acceptable terms within a few weeks."
Anthropic received a $4 billion investment from Amazon and, two years later, another $2 billion investment from Google.
After OpenAI received an additional $2 billion investment from Microsoft, Altman and another executive, Greg Brockman, went to Bill Gates' mansion on the shores of Lake Washington, outside Seattle. Although the Microsoft co-founder was no longer involved in the company's day-to-day operations, he still maintained regular contact with the company's executives.
During dinner, Gates told them that he was unsure if the large language model would work. He said that he would remain skeptical unless the technology could perform tasks that required critical thinking, such as passing a college introductory biology exam.
Five months later, on August 24, 2022, Altman and Brockman returned, accompanied by an OpenAI researcher named Chelsea Voss. Ms. Voss had been an international biology Olympiad winner in high school. Nadella and other Microsoft executives were also present.
In the living room outside Gates' house, a large digital display, the OpenAI team demonstrated a technology called GPT-4.
Brockman had the system take a multiple-choice test for an advanced biology exam, and Ms. Voss graded the answers. The first question involved polar molecules, a group of atoms with one end carrying a positive charge and the other end carrying a negative charge. The system correctly answered the question and explained its choice. "We only trained it to provide answers," Brockman said. "The conversational aspect miraculously emerged." In other words, it was doing things they hadn't really designed it to do.
Out of 60 questions, GPT-4 only answered one incorrectly.
Gates sat up straight, his eyes wide open. In 1980, when researchers showed him the graphical user interface that would become the foundation of modern personal computers, he had a similar reaction. He believed that GPT had that kind of revolutionary potential.
By October, Microsoft had already added this technology to its online services, including its Bing search engine. Two months later, OpenAI released its ChatGPT chatbot, now used by 100 million people every week.
OpenAI had defeated the effective altruists at Anthropic. Page's optimistic faction at Google quickly released their own chatbot, Bard, but it was widely considered to have lost in the competition with OpenAI. After the release of ChatGPT, Google's stock fell by 11%. Mr. Musk was nowhere to be found.
But this was just the beginning.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
