SAN FRANCISCO — Elon Musk celebrated his 44th birthday in July 2015, at a three-day party hosted by his wife in California wine country. It was family and friends only, with kids running around the exclusive resort in Napa Valley.
This was years before Twitter became X and Tesla had a profitable year. Musk and his then-wife, Talulah Riley, were a year away from calling it quits on their second marriage.
Larry Page, one of the guests at the party, was still the CEO of Google. And artificial intelligence had breached the public consciousness just a few years earlier, when it was used to identify cats on YouTube — with an accuracy of 16 percent.
AI was the big topic of conversation when Musk and Page sat near a campfire after dinner the first night. The two magnates had a friendship of more than a decade. But the tone on that cloudless night soon turned contentious as the two debated whether AI would ultimately elevate humanity or destroy it.
Page described his vision of a digital utopia. Eventually, humans would merge with AI machines, she said. One day there would be many types of intelligence competing for resources and the best would win.
If that happens, we're doomed, Musk responded. Machines will destroy humanity.
In frustration, Page insisted that his utopia must be pursued. Finally, she called Musk a “speciesist,” a person who favors humans over digital life forms of the future.
Eight years later, the discussion between the two seems to be prophetic. The question of whether AI will elevate the world or destroy it has framed an ongoing debate among Silicon Valley founders, chatbot users, academics and regulators over whether the technology should be controlled or unleashed.
That debate has pitted some of the richest men in the world against each other: Musk; Page; Mark Zuckerberg, from Meta; technology investor Peter Thiel; Satya Nadella, from Microsoft, and Sam Altman, from OpenAI. They have all fought for a piece of the business and for the power to transform it.
Musk and Page stopped speaking shortly after the party that summer. A few weeks later, Musk had dinner with Altman, who was then operating a technology incubator. That led to the creation of OpenAI. Backed by hundreds of millions of dollars from Musk and others, the lab promised to protect the world from Page's vision.
Thanks to its chatbot ChatGPT, OpenAI has introduced the world to the risks and potential of artificial intelligence. It is valued at more than $80 billion, two people familiar with the company's latest financing round revealed, although Musk and Altman have stopped speaking.
“There are disagreements, distrust and egos,” Altman declared. “The closer people are to being pointed in the same direction, the more contentious the disagreements are. You see this in religious sects and orders. There are bitter fights between the most united people.”
Last month, rebellious members of the company's board attempted to oust Altman because, they believed, they could no longer trust him to develop AI that would benefit humanity.
For five days, it looked like OpenAI was going to fall apart, until the board, pressured by big investors and employees who threatened to follow Altman, backed down.
The drama gave the world its first glimpse into the bitter disputes between those who will determine the future of AI. But years before, there was a little-publicized but fierce competition in Silicon Valley for control of the technology that is now rapidly transforming the world.
The New York Times spoke with more than 80 executives, scientists and entrepreneurs to tell that story of ambition, fear and money.
Five years before the party in Napa Valley, Demis Hassabis, a 34-year-old neuroscientist, attended a party at Thiel's home in San Francisco.
There, in Thiel's room, was a chess board. Hassabis had once been the second best player in the world at the under-14 level. “I was preparing for that meeting for a year,” he recalled. “I thought that would be a unique hook: I knew he loved chess.”
In 2010, Hassabis and two colleagues, living in Britain, were seeking money to develop “artificial general intelligence,” or AGI, a machine that would do anything the brain can do.
Thiel had become wealthy thanks to an initial investment in Facebook and his work with Musk in the early days of PayPal. He had developed a fascination with the singularity, a science fiction trope that describes the moment when intelligent technology can no longer be controlled by humanity.
Delighted with Hassabis at the party, Thiel invited the group to return the next day. The three made their sales pitch and Thiel and his venture capital firm soon agreed to invest £1.4 million (about $2.25 million) in their startup.
They named their company DeepMind. In the fall of 2010, they were building their dream machine. They believed that by understanding the risks, they were uniquely positioned to protect the world.
About two years later, Hassabis met Musk at a conference organized by Thiel's investment fund. Musk soon invested in DeepMind.
DeepMind hired researchers who specialize in neural networks, complex algorithms that identify patterns in large amounts of data. DeepMind developed a system that could learn to play classic Atari video games to illustrate what was possible. This caught Google's attention.
Hassabis had always told his employees that DeepMind would remain an independent company. He believed that was the best way to ensure that she did not become dangerous.
However, at the end of 2012, Google and Facebook were looking to buy it. The founders insisted on two conditions: no DeepMind technology could be used for military purposes and its AGI technology would have to be overseen by an independent board.
Google offered 650 million dollars. Facebook's Zuckerberg offered a larger payment, but did not accept the conditions. DeepMind was sold to Google.
When Musk invested in DeepMind, he broke his own informal rule — that he wouldn't invest in a company he didn't run. The disadvantages were evident a month or so after his discussion with Page. When Google bought DeepMind, Musk was out. Three Google executives were now in control: Page; Sergey Brin, co-founder of Google and Tesla investor; and Eric Schmidt, the president of Google.
Eight months later, DeepMind had a breakthrough: a machine called AlphaGo beat one of the world's best players in the ancient game of Go. The game, broadcast via internet streaming, was watched by 200 million people. Most researchers had assumed it would take AI another 10 years to achieve this.
Convinced that Page's optimistic view of AI was wrong, and furious at his loss of DeepMind, Musk developed his own lab.
OpenAI was founded in late 2015. In late 2017, it hatched a plan to wrest control from Altman and the other founders, four people familiar with the matter said.
When others resisted, Musk resigned and said he would focus on his own AI work at Tesla.
Suddenly, OpenAI needed new funding. Altman ran into Satya Nadella, the CEO of Microsoft, at a conference. A partnership seemed natural. The agreement was finalized in 2019.
After OpenAI received another $2 billion from Microsoft, Altman and another executive, Greg Brockman, visited Bill Gates at his sprawling mansion outside Seattle.
Over dinner, Gates told them that he doubted that large language models could work. He noted that he would remain skeptical until the technology performed a task that required critical thinking—passing an AP exam in biology, for example.
Five months later, on August 24, 2022, Altman and Brockman returned and brought with them an OpenAI researcher named Chelsea Voss, who had been a medalist in an international biology Olympiad when she was in high school.
On a huge digital screen placed outside Gates' room, the OpenAI team presented a technology called GPT-4. Brockman gave the system a multiple-choice advanced biology test and Voss graded the answers. There were 60 questions. GPT-4 got only one answer wrong.
By October of that year, Microsoft was adding the technology to its online services. And two months later, OpenAI launched its ChatGPT chatbot, which is now used by 100 million people every week.
Page's optimists at Google were quick to launch their own chatbot, Bard, but it was widely perceived that they had lost the race to OpenAI. Three months after the launch of ChatGPT, Google shares fell 11 percent. Musk was nowhere to be seen.
But it was only the beginning.
CADE METZ, KAREN WEISE, NICO GRANT AND MIKE ISAAC. THE NEW YORK TIMES
BBC-NEWS-SRC: https://www.nytimes.com/2023/12/03/technology/ai-openai-musk-page-altman.html, IMPORTING DATE: 2023-12-14 19:40:07
#Artificial #Intelligence #elevate #humanity #destroy #eternal #debate