On 17 November local time, the official announcement of OpenAI, an American artificial intelligence research company, announced that Sam Altman will step down as CEO and quit the board of directors.
The announcement was quite harshly worded: “Mr Altman’s departure is the result of a considered decision by the Board of Directors, which believes that Mr Altman has not been sufficiently forthcoming in his communications with the Board of Directors and has impeded the Board’s ability to fulfil its responsibilities. The Board no longer believes he is capable of continuing to lead OpenAI.”
From a post by Greg, OpenAI’s chairman and president, the firing went like this: on Thursday night, Altman received a text message from Ilya, the chief scientist, about a meeting on Friday at noon. At noon on Friday, Altman accessed the meeting, the rest of the board, except Greg, were there, and Altman was told he was fired. At 12:19 p.m. Greg receives a call from Ilya, and at 12:23 p.m. Greg accesses the meeting and is told that he has been removed from the board of directors and continues to stay in management, and that Altman has been fired.
At the end of November last year, with the shiny introduction of ChatGPT, a wave of artificial intelligence swept the world, and Altman was honoured as the “father of ChatGPT”. A year later, he was fired, quite dramatic.
On November 6, Altman announced a series of product updates to developers and ChatGPT users around the world at OpenAI’s inaugural developer conference, including the ability to quickly create custom ChatGPT GPTs to realise everyone’s dream of owning a big model, and said he would compete with the likes of Microsoft and Google.
On Thursday, Altman also spoke at APEC, expressing bullishness on AI technology, arguing that the technology doesn’t need major regulation in the short term, and even claiming, “As a species, humanity is now on a path to self-destruction, and AI may be the means to stop it.”
And then suddenly he was “out of office” on Friday.
At 5:46 a.m. on November 18, Altman responded with a message on the X platform, “I loved my time at OpenAI, it changed me and hopefully it will change the world. Most importantly I have worked with so many talented people. I’ll explain more about what’s next later.”
OpenAI CTO Mira Murati, formerly CTO, has been named interim CEO.At the same time, company president Greg Brockman announced that he would be stepping down as chairman of the board of directors, but that he would continue to hold a position at the company, reporting directly to the CEO.
The board described interim CEO Mulati as technically well-rounded and particularly experienced in AI management and policy implementation. Previous reports have called her “a dynamic and conciliatory figure in AI” who has welcomed government regulation.
On Friday, she told employees at a company-wide meeting that Microsoft’s relationship with OpenAI remains stable following Altman’s departure. By the news of Altman’s departure, Microsoft shares late in the session once “diving” fell more than 2%, closing down 1.68%. Microsoft issued a statement that the company has a long-term relationship with OpenAI, abide by the investment commitment to OpenAI.
As you can see, Altman was ousted from the board for at least two reasons:
One is that he disagrees with the board on the issue of how to handle the relationship with Microsoft.
The second is a disagreement with the board on the question of whether regulation of AI should be prevented in the short term.
I. Begging for a package or independence?
Microsoft is by far the largest investment in OpenAI, after investing $1 billion in 2019, and another $10 billion in early 2023.
Microsoft and OpenAI were once called a “romantic alliance” in the tech world. However, if you look closely, you will find that the shrewd Microsoft has not lost its mind in the face of “romance”. 2019 investment of $1 billion in OpenAI, the conditions are very favourable to Microsoft, buy out the exclusive license of the basic technology of the GPT-3, most of the technology of OpenAI is preferentially licensed to the Microsoft products, and Microsoft has become the exclusive cloud provider of OpenAI, and Microsoft has become the exclusive cloud provider of OpenAI. OpenAI’s exclusive cloud provider, meaninga large portion of the amount of Microsoft’s investment that OpenAI received was to be paid to Microsoft as a cloud usage fee.
Last November, OpenAI released the ChatGPT chatbot, which took the world by storm within a few weeks. At the beginning of February this year, ChatGPT turned on a wild ride, and a new global AI race was triggered, with giants competing for the top spot. Microsoft immediately ChatGPT access to Microsoft cloud services, the launch of GPT-3, GPT-4 version of the Office Family Bucket and a series of new products, especially Microsoft ChatGPT into its search engine to launch a new version of the Bing artificial intelligence search engine (Bing Chat), industry insiders believe that this is likely to help Microsoft arm in arm, in the search market to conquer the city.
Every time a new technology emerges, it’s time for a major industry reshuffle. Although in the global search market in October 2023, Google occupied 91.55% of the market share, while Bing was only 3.11%, which is still a slight decline from last year, but artificial intelligence, a new technological wave hit over, the uncertainty of the game will be greatly increased, which is a huge challenge for the leader Google, while for the follower of Microsoft is a huge opportunity.
Microsoft obviously does not want OpenAI to provide products and services for other customers, the most favourable situation for it is OpenAI only as an enabler behind it, do not come out to the foreground to receive customers. At the same time, Microsoft has other enablers, partners, to ensure flexibility. It’s like a romantic relationship where both parties want to have many lovers and the other party has only one lover of their own.
A Microsoft internal document in March this year shows that in order to persuade customers to choose their own Azure OpenAI service, Microsoft secretly ordered its sales staff to “pull the pedal” OpenAI. when Meta re-released the open source large language model LIama 2, Microsoft announced that it had become the first partner of Llama 2.
Altman is obviously not willing to accept such a “adopted” role, and is also frequently contacting new “gold masters”.
Just when he was confident that he would survive on his own, OpenAI’s board of directors staged a “coup” and drove him out of the company overnight.
II. A for-profit company controlled by a non-profit organisation
While Microsoft has made a large investment in OpenAI, it is investing in OpenAI Global, which was founded in 2019 and is “fully controlled” by the OpenAI non-profit organisation, which was incorporated in Delaware in 2015.
OpenAI Global is “permitted to earn and distribute profits” subject to the mission espoused by the parent organisation, OpenAI, which has always emphasised Microsoft’s “acceptance of our cap-and-trade offer, and the fact that we are keeping the AGI technology and governance retained in the hands of the non-profit and indeed all of humanity”.
In an interview in January 2023, Altman also said, “The future I would like to see is one where access to AI is super-democratised, and there are multiple AGIs in the world that can help people form multiple perspectives without making anyone too powerful.”
Under Microsoft’s investment agreement with OpenAI, Microsoft’s share of the profits gained by OpenAI Global is limited to “gains made prior to the realisation of AGI (Artificial General Intelligence)” as set out in OpenAI’s charter.
“AGI” is defined as “a highly autonomous system capable of outperforming humans in the most economically valuable tasks.” So who decides whether or not to “realise AGI”?
The six members of OpenAI’s board of directors, known as the “Open AI Six”: Chairman and President Greg Brockman, Chief Scientist Ilya Sutskever, Chief Executive Officer Sam Altman, and three independent directors who are not OpenAI employees, Adam D’Angelo, Tyler D’Angelo, and Tyler D’Angelo. Angelo, Tasha McCauley and Helen Toner.
If they believe that AGI has been “realised”, then Microsoft no longer has the right to make a profit. While it is rare for a non-profit board to decide whether shareholders of a for-profit company will receive a profit, OpenAI is unique in that it was founded by Altman and others as a non-profit organisation with a mission to ensure safe general artificial intelligence (GAI) for the benefit of mankind.
They saw the non-profit board as focused on benefiting all of humanity, while the for-profit board served investors. They have also endeavoured to build in another layer of security, which is to keep the board majority independent, i.e., a majority of the members do not hold equity in OpenAI Global.
It is argued that there are indeed some non-profit organisations that own shares in for-profit organisations, most notably the Hershey Trust. But they are in full control of the for-profit companies under them, with no opposition from minority shareholders. In the case of OpenAI, Microsoft’s for-profit interests could directly conflict with the nonprofit interests of the controlling entity.
Once OpenAI achieves its proposed AGI mission, Microsoft fears it will be cut out of the picture. But at last week’s OpenAI Developer Day event, Altman assured Microsoft CEO Satya Nadella: “I highly appreciate the partnership between the two parties on a technical level …… and look forward to working together to achieve the AGI mission.”
In a new interview with Altman in the Financial Times on 13 November, the OpenAI boss said that the company’s partnership with Microsoft is “working well” and that he expects “we’ll be able to raise more money over time.”
III. Idealist Altman
In 2005, Altman dropped out of school with two of his classmates to start a business, developing Loopt, a social media software that enabled them to share their location with their friends, and the famous startup incubator Y Combinator invested in the software.
Three years later, Steve Jobs invited Altman to attend the Apple mobile phone launch event, and the software entered the Apple APP shop. Since then, Altman has developed some more software, all of which were popular, and after operating for a while, sold it. in 2011, he became a partner of Y Combinator.
In the summer of 2015, at a Silicon Valley hotel, Altman hosted a private party to which he invited several industry luminaries, including Tesla’s Musk as well as Musk’s former Paypal partners Peter Thiel and Reid Hoffman.
At the time, they saw that Silicon Valley giant Google had acquired AI company Deep Mind for $400 million, and agreed that if Google and Deep Mind were successful, they would likely have a monopoly on AI technology. They couldn’t let that happen.
On 11 December that year, OpenAI was announced, with funding from Altman, Musk, Hoffman, Thiel and others, with a mission to ensure safe general artificial intelligence (AGI) for the benefit of humanity. At the time of its inception the nature was that of a non-profit organisation, which caught the attention of the industry.
Deep Mind released the sensational Alpha Go the following year, while Open AI was still in its infancy. Altman resigned from Y Combinator in 2019 to become CEO of OpenAI so he could focus on growing OpenAI.
It soon became clear to Altman that achieving the mission would require significant funding, which was hampered by the nature of a not-for-profit organisation, and in March 2019 OpenAI announced the creation of a for-profit entity to ensure that it would be able to raise sufficient funding while retaining the mission, governance and oversight of a not-for-profit organisation. The reason for this “nonprofit + for-profit corporation” structure is that “no existing legal structure that we know of strikes the right balance.”
It is indeed a paradox that in order to achieve his mission of “ensuring safe general artificial intelligence (AGI) for the benefit of mankind”, he has to raise huge sums of money, which needs to be returned, and a non-profit organisation can’t achieve that goal. So another for-profit entity, a limited liability company, had to be formed.
Altman also made an unusual decision:He would not take an equity stake in the company. Having invested in several highly successful tech startups himself, he was already very wealthy and didn’t need the money. Avoiding any equity would help him stay in line with his original mission.
This decision, however, actually turned off some of OpenAI’s potential investors, who suspected that by not taking an equity stake, Altman was not seeing a return on the project. Still, in July 2019, OpenAI received a $1 billion investment from Microsoft.
IV. Is being kicked out of the company you founded a sign, an honour?
As soon as the news of Altman’s dismissal broke, Musk’s social media platform “X” posted a “link to the job application” on its official account, with a “just in case anyone needs it”. Mockery with sympathy, which is very consistent with Musk’s characteristics, perhaps he is throwing an olive branch to Ottoman.
In early 2018, Musk told Altman that he thought OpenAI had fallen a long way behind Google, and offered to take control and run OpenAI himself.Back then, this had happened at Tesla, and Musk took over and was really successful, and he may have wanted to repeat the Tesla experience.
However, this time he didn’t get control, as Altman and the other founders unanimously rejected the offer. So on 20 February 2018, Musk announced that he was stepping down from OpenAI’s board of directors and would no longer be involved in its affairs in any way, citing a conflict of interest between Tesla’s development of Autopilot technology and OpenAI.
At the end of November 2022, ChatGPT exploded in popularity as soon as it was launched, and Musk was furious.On 17 February 2023, he tweeted, “OpenAI was created under the banner of open source, and that’s why I named it ‘Open’ ( ‘Open’) AI, originally a non-profit to be a force against Google, and now it’s become a closed-source, Microsoft-controlled, profit-maximising corporation.”
On 15 March, Musk tweeted again, “I’m still confused as to how a non-profit organisation that I donated about $100 million to in the first place has now become a for-profit company’s with a $30 billion market cap. If this is legit, why isn’t everyone following suit?”
Recently, OpenAI was in talks to sell its current employees’ shares at a valuation of $90 billion, with valuations skyrocketing over the course of the year.
Musk is understandably angry that non-profit organisations suddenly become for-profit companies, and that $100 million, if not a donation to a non-profit organisation, but an investment in a for-profit company, will still get him shares; however, that shouldn’t be the main reason for his anger.
A hundred million dollars is a piece of cake for him, however, he believes that Altman betrayed the original intent, purpose and mission of OpenAI when he helped him found it, which was supposed to be against Google, a potential monopoly in the field of AI, and ended up fuelling a real-life monopoly, Microsoft.
Now that Altman has been kicked out of OpenAI, it proves that Musk was wrong about him and that he and Microsoft are not in cahoots. Instead, it’s just the right time to go to Musk to continue working on AI against the monopoly that is Microsoft.
On 12 July this year, Musk announced in a tweet the establishment of an artificial intelligence company xAI, the company’s goal is to understand the true nature of the universe. 5 November, Musk’s xAI team released its first AI large model product - Grok. according to the introduction, Grok understands the world in real time through the X platform, and is also able to answer questions that are rejected by most of the other AI systems. other AI systems reject tough questions.
Musk will be happy to “take in” Altman, as he himself has twice been kicked out of companies he founded by the board of directors, a traumatic experience that will last a lifetime, and he was kicked out for reasons similar to Altman’s - he wanted to pursue his dream of building an independent, great company, not a third-rate one that was underwritten by the board of directors.
The other reason of course was his pugnacious management style. Either way, he should be easier to empathise with Ultraman.
Musk’s first pot of money came from Zip2, the online yellow pages company he founded. He wanted to buy the domain name “city.com” to compete head-to-head with Yahoo and America Online, but the board of directors preferred to relegate the product to the status of a no-name supplier to newspaper conglomerates. Musk resisted, and the board threw him out, reducing his power at the company and driving him out in disguise.
He was initially prepared to fight to the end, then interim CEO Prodion advised him, “This is your first company, let’s find an acquirer and make some money so you can do a second, third, and fourth company.”
Four years after Musk and his brother Kimball founded Zip2 in January 1999, Compaq bought it for $307 million in cash. 27-year-old Musk received $22 million.
He then started a second company, online payments company X.com, which later merged with Confinity, an online payments company founded by Thiel, Levchin and others, to launch the PayPal service. Till and the others took advantage of his honeymoon to stage a “coup d’état” and kick him out of the company.
Musk felt like he’d been stabbed in the back. “I’m so saddened by this that I can’t even begin to describe it,” he wrote in an email. “I’ve worked my ass off for this company, I’ve got almost all the cash on the books from Zip2, my marriage is on the line, and they say I’m full of shit, and they’re saying I’m full of shit. And they say I’m so full of evil that they don’t even give me a chance to complain.”
PayPal went public in 2002 and was acquired by eBay for $1.5bn in July that year. Musk got about $250 million in return. But it could have grown into a multi-trillion dollar company, and Musk started it back then with a vision of social media + a mega-bank, akin to what we’ve come to call WeChat here. His insistence on acquiring Twitter was to realise this early dream.
He said, “This is the mission that Twitter could be fulfilling in the future. You can create what I think X.com should be if you combine social media with a payment platform.”
Later Musk told Inc. magazine:”Great things are never born in the hands of venture capitalists and professional managers. They don’t have the creativity or the insight.”
Twice in three years, Musk was ousted from the company he founded, a far more harrowing experience than Steve Jobs was ousted from Apple back in the day. Apple deteriorated after Jobs left and finally had to bring Jobs back again to save the day.
I don’t know which script God gave to Ultraman, no matter what, people who have contributed to the development of human technology and progress of civilisation always deserve everyone’s respect, bless him.