Shulou(Shulou.com)11/24 Report--

[director Xin Zhiyuan] the OpenAI palace fighting drama is coming to an end. Altman and Brockman join Microsoft hand in hand. As the only remaining core member, Ilya, what are the behind-the-scenes stories?

The OpenAI palace drama has finally reached the end of its first season:

Sam Altman finally left OpenAI with a visitor's card, and Greg Brockman drove out of California to Washington State to join Microsoft.

At present, only Ilya Sutskever, the chief scientist, is still on the board of OpenAI.

The Russian-born AI researcher suddenly went from behind the scenes to the foreground and became the absolute core of OpenAI.

Where will the AI scientist, whom Musk calls "OpenAI's hardest and most important recruitment", lead OpenAI?

Now there are only four of the top six King Kong on the board of OpenAI. What changes have taken place in the board of directors along the way?

In order to recruit Ilya, Musk did not hesitate to break with Larry Page. Ilya Sutskever was born in Nizhny Novgorod, Russia, a city known as Gorky in Soviet times. He emigrated to Israel with his family at the age of five and grew up in Jerusalem.

Sutskever studied at the Israel Open University from 2000 to 2002. After that, he emigrated to Canada with his family and transferred to the University of Toronto.

He received a Bachelor of Science in Mathematics from the University of Toronto in 2005, a Master of Science in Mathematics in 2007 and a PhD in computer Science in 2013. His doctoral supervisor is Geoffrey Hinton.

AlexNet, which he developed with Hinton, crushed other big companies in the ImageNet competition, improving recognition accuracy by 60 per cent.

From November to December 2012, Sutskever did postdoctoral research with Wu Enda at Stanford University for about two months. Then, at the call of Hinton, he returned to the University of Toronto and joined Hinton's new research company, DNNResearch.

Google had the last laugh in the well-known auction of "Baidu, Google, DeepMind, Microsoft Business War" and successfully recruited Hinton and Ilya.

After joining Google Brain, Ilya worked with two other colleagues to create the sequence-to-sequence Learning algorithm and participated in the development of TensorFlow.

At the end of 2015, he was poached to OpenAI by Musk to become co-founder and chief scientist.

Musk said in a podcast on Fridman on Nov. 9 that his relationship with the Google founder could have been repaired because he broke up when he dug Ilya into OpenAI.

"[Google DeepMind] Hassabis was on one side and I was on the other side [fighting for Ilya]," Musk said. It was one of the toughest hiring battles I've ever been through, but it was really the key to OpenAI's success. "

Musk added, "Sutskever is not only smart, but also a kind-hearted good person (a good human-smart, good heart)."

The man at the center of OpenAI power, Ilya, who has always avoided the media spotlight, gave a lengthy interview with the Massachusetts Institute of Technology Review (MIT Technology Review) last month.

The Israeli-Canadian told the magazine that his new research focuses on how to prevent super artificial intelligence from getting out of control, although it doesn't exist yet.

Earlier this year, Hinton, Ilya's mentor and godfather of artificial intelligence, left Google with the same goal.

Hinton warns that the arms race of artificial intelligence companies is causing danger to mankind, such as OpenAI's ChatGPT, and it is difficult for us to prevent bad guys from using it to do bad things.

In 2012, Hinton and two of his graduate students, including Ilya, developed the famous AlexNet, demonstrating the power of neural networks in pattern recognition-much better than widely recognized.

Later, Google acquired DNNresearch, Hinton's company, and hired Ilya. While working at Google, Ilya helped prove that AlexNet's pattern recognition ability in images applies to words and sentences as well.

After being poached by Musk to OpenAI, Ilya played a key role in developing large language models (LLM), including GPT-2, GPT-3, and the text-to-image model DALL E.

ChatGPT, released at the end of last year, gained 100 million users in less than two months, setting off the current artificial intelligence boom.

Ilya says ChatGPT makes people see the possibility of artificial intelligence, although sometimes it returns incorrect results and disappoints people.

Ilya's recent focus is on the potential dangers of artificial intelligence, which he believes could emerge within a decade.

The super artificial intelligence mentioned by Ilya is not just the general artificial intelligence (AGI) that can compete with human beings, it will be smarter and more powerful than human beings.

The debate about the safety of artificial intelligence is the trigger for the change of OpenAI leadership.

There are differences between Ilya and Altman on how to quickly commercialize the production of artificial intelligence products and what steps to take.

Ilya said in an interview that it is important that the super intelligence built by anyone can not be defected.

To this end, Ilya wants to guide artificial intelligence systems to be consistent with people's expected goals or moral principles, rather than allowing it to pursue unexpected goals.

In July, Ilya and colleague Jan Leike wrote an OpenAI announcement about a project on super-intelligent alignment.

They warn that while super intelligence can help solve many of the world's most important problems, it can also be very dangerous and can lead to human disempowerment and even extinction.

How much OpenAI's board knows that Altman is out, Brockman resigned, and OpenAI's show is as hot as its own developer conference-it's more important than development.

In order to eat melons better, the editor is here to help you sort out a timeline for changes in the board of directors of OpenAI.

Is everything today already foreshadowing in the past? Let's find out.

OpenAI was founded on December 11, 2015, and the board is co-chaired by Elon Musk and Sam Altman.

On December 31, 2016, the Board joined Chris Clark (as COO) and Jonathan Levy (Secretary, Treasurer).

In March 2017, Open Philanthropy donated $30 million to OpenAI, and its founder Holden Karnofsky joined OpenAI's board of directors.

On December 31, 2017, public information showed that members of the board of directors added Greg Brockman (a good friend with Altman) and Ilya Sutskever (which was poached by Boss Ma above).

On February 20, 2018, Musk was removed from the board. Officials say Musk left to avoid potential conflict, but reporters reported that differences in the leadership were the main reason.

Reid Hoffman, founder of LinkedIn and partner of Greylock, joined the board in March 2018.

Adam D'Angelo, CEO of Quora and former CTO of Facebook, joined the board in April 2018.

In September 2018, Sue Yoon joined the board. Then, in November 2019, she left the board.

By December 31, 2020, official public information showed that the members of the board of directors were Ilya Sutskever, Greg Brockman, Sam Altman, Reid Hoffman, Shivon Zilis, Holden Karnofsky, Adam D'Angelo and Tasha McCauley.

Will Hurd, a Republican member of the House of Representatives and a former CIA agent, joined the board on May 3, 2021.

On September 8, 2021, Helen Toner, director of the Center for Security and emerging Technologies at Georgetown University and former director of Holden Karnofsky Open Philanthropy, joined the board.

Holden Karnofsky resigned from the board in the fall of 2021 because his wife, Daniela Amodei, was working with her brother Dario Amodei to start OpenAI's main competitor, Anthropic.

Given that Helen and Holden have opened up philanthropic links, and that Holden's board seat seems to be permanent, Helen seems to have been chosen by Holden to replace him.

In the days that followed, some people left the board for a variety of reasons, leaving only the famous "six King Kong": Ilya Sutskever, Greg Brockman, Sam Altman, Adam D'Angelo, Tasha McCauley and Helen Toner.

Reference:

Https://fortune.com/2023/11/18/who-is-ilya-sutskever-openai-leadership-shakeup-ai-safety/

This article comes from the official account of Wechat: Xin Zhiyuan (ID:AI_era)

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.