Ultramans Return OpenAI Negotiations Stalled, Replacing Board Members as Major Obstacle How Will It End? How to View this Game?

Altman’s Return and Changes to the OpenAI Board

Let’s start with a funny picture.

This is a selfie taken by an OpenAI employee today (two days after Sam Altman was fired), showing their former CEO taking a selfie.

The selfie was taken at the OpenAI office.

So, what’s this selfie all about?

Sam Altman posted a tweet with a selfie of himself holding an OpenAI visitor pass.

Two days ago, he was still the CEO, but today he has become a visitor.

His tweet said, “first and last time i ever wear one of these.”

I think there are two meanings to this.

The first time he wore a visitor pass to enter OpenAI because he used to be an employee of OpenAI and naturally did not need one.

The last time he intends to resolve this issue, either to continue as the CEO and return like the Dragon King, or he quits.

In fact, Altman’s return began when he was fired. It was not only his own desire to come back, but also because OpenAI kicked out a board member that even the main investor Microsoft was unaware of.

This is indeed a bit excessive.

It feels like acting first and reporting later, at the very least they should have notified in advance.

The core person who led Altman’s return is Satya Nadella, the CEO of Microsoft.

A couple of days ago at OpenAI DevDay, there was still close interaction with Altman.

It seems that OpenAI’s board couldn’t withstand the pressure from the main investor Microsoft to kick Altman out.

But from the news, the condition for Altman’s return is the dissolution of the current OpenAI board.

This includes IIya, the chief scientist.

For those who don’t know him, you can think of him as Oppenheimer.

Oppenheimer led the development of the atomic bomb in the United States, but later he raised the banner against nuclear weapons.

Feels similar to IIya. He initially led the development of GPT, but by the fourth or even the fifth generation, he began to feel that the development was too fast and that AI safety should be prioritized.

However, based on the experience of atomic bomb development, slowing down is impossible because others are also in hot pursuit.

Only when you are significantly ahead of others can there be a balance.

On the other hand, Sam Altman is actually more like a computer background professional businessman. He is someone who sees technology as a means to an end.

You can see this from the companies he has worked for, all over the place.

He worked on Reddit, the foreign version of Tieba (Baidu Post); Y Combinator, a venture capital company; Worldcoin, a cryptocurrency.

These are not the only ones. On his blog, he is even involved in nuclear fusion.

It can be said that his dedication to technology cannot be seen.

Just like Musk, who has done a lot of things, but it seems that his vision has never been about human society.

The so-called technology is nothing more than a stepping stone to achieve personal ideals.

I think it is highly likely that Sam will return, and IIya will also stay. The board members other than IIya will be kicked out, a new board will be formed, and Microsoft may add someone.

Otherwise, with an investment of $10 billion, not even having a board is a blatant injustice.

“OpenAI Personnel Adjustment and Development”

As a staunch advocate for the timely arrival of AGI, I don’t actually hope for an overwhelming victory for either Sam or Ilya. As long as they sit down and talk, progress can be made.

First, a few personal expectations/predictions (just speculations, could be completely wrong):

1. Sam and Greg return to OpenAI.
2. The existing independent directors on the board are all replaced, with Microsoft gaining at least one seat.
3. Ilya stays at OpenAI for the time being, but the future is uncertain.
4. Although it’s unlikely to happen and Musk is quiet, it would be interesting if he gets involved too. It would have a bit of an Avengers vibe.

Addendum: Musk seems to be supporting Ilya:

Sam has returned to the OpenAI office, although as a visitor. He said, “This is his first and last time wearing a visitor badge.” It’s unclear whether he is confident about returning or if he means to leave for good…

Visitor 04

(I’m curious about who visitors 01-03 are)

There is a greater hope for Sam’s return, as OpenAI faces immense pressure internally and externally.

Internally, on one hand, some members of the management are requesting the board to bring Sam back. After Sam’s tweet yesterday, even the new CEO Mira came out in support, and COO Brad Lightcap expressed optimism internally about Sam’s return. On the other hand, a large number of technical staff have also expressed their stance and may resign (some have already filed their resignations).

Externally, on one hand, investors such as Microsoft, Sequoia, and Tiger are pushing for Sam’s return. On the other hand, AI companies are actively poaching talent. It is said that competitors like Cohere and Adept are making efforts to recruit OpenAI employees, and Google DeepMind has previously received more resumes from OpenAI employees.

If the situation is not resolved soon, OpenAI is already facing a practical disintegration.

---

The original triumvirate at OpenAI had a good division of labor. Sam (the manager) was responsible for fundraising, publicity, and management; Greg (the engineer) was responsible for engineering management, deployment, and infrastructure; Ilya (the scientist) was responsible for technical vision and direction. The three of them together facilitated the development of ChatGPT and GPT-4.

I hope that through this negotiation, a more reasonable balance can be achieved between AI safety and development, without wasting ChatGPT’s leading position. At the same time, efforts should be made to address the company’s product-level issues, allowing technical personnel some time to fix the bugs in ChatGPT and the API, clarify the capacity of computational resources, and contemplate the next steps for ChatGPT and GPT-5.

Internal Negotiations: Ultraman vs. Board of Directors

There is a bloodless negotiation taking place within OpenAI, involving the company’s founder Ultraman and whether he can regain his position as Chief Executive Officer (CEO) at OpenAI, as well as the potential changes to the board of directors.

According to insiders, Ultraman himself is open to returning to OpenAI but hopes to see a change in management, including the removal of current board members. Additionally, Ultraman has requested that the company issue a statement to absolve him of any wrongdoing.

Another source claims that the leadership team at OpenAI is also pushing for the resignation of board members and for Ultraman to return, including interim CEO Mira Murati, Chief Strategy Officer Jason Kwon, and Chief Operating Officer Brad Lightcap.

Last Friday, the board members led by OpenAI’s Chief Scientist Ilya Sutzkever suddenly announced Ultraman’s dismissal and stripped him of his board position. At the time, the board claimed that Ultraman had not been forthcoming in communication with them.

Subsequently, OpenAI President Greg Brokman was also removed from the board, and he announced his resignation in protest of the events of that day.

Ultraman, Sutskever, and Brokman, who are both co-founders and core team members of OpenAI, along with interim CEO Murati, hold the reins of this AI giant. Sutskever’s action of pushing out his two old partners undoubtedly brings a sense of “palace intrigue” to the whole affair, and has captivated the tech community.

Rivalry between Research and Development

According to Brokman’s account, last Thursday evening, Sutskever received a message from Ultron requesting a meeting the following day. Then, on Friday noon, in an online meeting attended by everyone except Brokman, the board suddenly announced that Ultron would be terminated and informed Ultron that this news would spread quickly.

Shortly after, Sutskever convened another board meeting without Ultron present, and Brokman learned that he would be expelled from the board but would retain his position in the company. Brokman also claims that all other OpenAI executives, except for the interim CEO Murati who was put in place, were unaware of this personnel change.

What differentiates the OpenAI board from traditional companies is that its purpose is not for profit, but rather, its mission is to create a widely beneficial general artificial intelligence.

Currently, in addition to Sutskever, the OpenAI board consists of Adam D’Angelo, CEO of Quora, Tasha McCauley, former CEO of GeoSim Systems, and Helen Toner, Director of Strategy at Georgetown’s Center for Security and Emerging Technology.

Observers interpret Ultron’s dismissal as a joint effort by Sutskever and external board members to seize power. The underlying cause is believed to be Ultron’s excessive ambition, which made the board uneasy and went against Sutskever’s stance on AI safety. Technologists comment that this is a typical case of discord between the research and product departments.

Mass Resignation and Palace Intrigue

The firing of Ultraman was just a momentary triumph. It was followed by a backlash from investors and internal technical staff, which thoroughly embarrassed the board led by Sutskever.

Several investors, including Microsoft, reached out to OpenAI, hoping to bring back Ultraman as CEO. It was rumored that Microsoft offered cloud computing services in exchange for Ultraman’s leadership position.

Even top AI scientists within OpenAI voiced their grievances for Ultraman, and some even threatened to collectively resign. According to insiders, after Ultraman and Brokman left, a group of senior researchers also resigned. More people are considering resigning, and a wave of departures from OpenAI is imminent.

Ironically, another co-founder of OpenAI urgently placed an advertisement on Musk’s X platform, hanging a recruitment link for his own company, and encouraging ambitious individuals to consider xAI.

Under this pressure, the OpenAI board began considering recalling Ultraman and Brokman, while the employees continued to threaten. The company had a limited amount of time to come up with a reasonable solution, otherwise it would face a flood of resignation letters.

The Most Crucial Puzzle

An uninterrupted series of reversals, comparable to “The Legend of Zhen Huan”. However, the perplexing question is why Sutskever, together with the external board members, expelled Ultraman and Brokman.

Some people have speculated that if it were just a power struggle within the company, why would this personnel change be so sudden and unprepared, to the extent that Sutskever and others are now caught in the crossfire, and are likely to be shown the door.

Or perhaps the reason lies not in the complexity of human society, but rather in some desperate and even insane problems with the artificial intelligence products that have emerged, causing Sutskever despair?

Considering that Sutskever has always been responsible for limiting the harm of artificial intelligence at OpenAI and frequently voices alarming statements such as “AI leads to the extinction of humanity”, coupled with his teacher, the father of artificial intelligence Geoffrey Hinton, always emphasizing “regretting the development of AI”, maybe there are more complex reasons behind his decision to expel Ultraman.

Up until now, the OpenAI board has been unable to explain what exactly is meant by the “lack of candor” mentioned in the dismissal statement, leaving the tech community speculating about this cryptic melon. Musk has even stated that the OpenAI board has an obligation to disclose the reason for Ultraman’s dismissal, as it could be strongly related to the artificial intelligence models.

This is the most crucial puzzle in the whole story. Whether it is a case of the board overestimating their own cleverness and falling into their own trap, or a more cyberpunk-esque plot, everything may depend on the motivations behind the OpenAI board’s decision to dismiss Ultraman.

Source: Cailian Press

OpenAI’s important decision and anticipated progress

Recording my thoughts on OpenAI: The moment for AGI, the “Oberman moment,” has come, and no matter how we choose, we cannot stop the course of history. But it is crucial to consider when that day will come and how humanity will face it. I respect Ilya for bringing up the safety risks of AGI, and no matter what, making society and the general public aware of what we are about to face. He is a warrior. Although it has caused quite a stir, he has indeed shown the public the pressing safety concerns within his capabilities.

I am extremely excited for Sam Altman’s return. With his strong fundraising ability, clear vision, and execution power, there are very few individuals like him who are capable of driving human progress. For OpenAI, it is not yet time to seek individuals who prioritize stability and securing their positions. The window of opportunity is not broad, and Sam is irreplaceable.

Essentially, I believe we are facing an unprecedented choice: whether to accelerate and confront the risk of a complete social upheaval or to slow down and temporarily solve the current AI safety dilemma. The moral dilemma of this route selection is right in front of us.

As a bystander, I am very curious about Sam Altman’s next move. GPT-5, a general artificial intelligence, is already within reach. Based on this, evaluating valuation and raising funds is an inevitable step in continuing research and realizing the vision for Sam and Greg.

Ilya is also an exceptional talent, and this incident has caused a rift. Whether Sam chooses to return and how to retain talent is a decision to be made: whether to keep himself, Ilya, or to have a generous coexistence to jointly drive the development of GPT-5 and safe AGI. This significant decision ultimately determines which direction OpenAI will take.

It is a difficult choice, and I am very much anticipating the progress that follows.

Empty String

Looking forward to this series, haha.

Internal Technical Differences at OpenAI

Personnel Struggles?

Yesterday’s dismissal was actually driven by OpenAI’s Chief scientist, Ilya. As scientists, we all have lofty ideals but tend to overlook the reality of the situation. Ilya, as a scientist, believes that most of the computing power should be dedicated to scientific research. However, as the CEO, Ultraman believes that if the computing power is not used to serve clients, where will the revenue and financing come from to build a larger computing platform? Such technical differences are very normal in technology companies.