In a previous post, GreenSI noticed that the news of AI was always very rhythmic every week. This week will not have failed to respect this rule. While Microsoft had just finished his Ignite conference with major announcements about Copilot, that Xavier Niel was throwing Kyutai, a French research lab to counter OpenAI, the OpenAI board dismisses Sam Altman from the post of CEO, emblematic figure who has helped to raise the company to the top of innovation in artificial intelligence.
So as not to hide anything from you when GreenSI saw this ad in his eve, we believed the hoax. But the information is very real and OpenAI has published a press release quite quickly to appoint a transitional CEO, the current CTO Mira Murati. We can also consult this ZDNet article to follow this event. This post focuses on the analysis of the impact of this event.
GreenSI has no doubt about the ability to bounce back from Sam, whose phone should not stop ringing this weekend, but this “transition” as OpenAI talks about, is far from trivial and raises a multitude of questions about the future of OpenAI, but also on the evolution of the generative AI landscape.
OpenAI, known for its revolutionary advances such as GPT and DALL-E, has played a central role in the democratization and innovation in AI over the past 12 months. Even if initially there were quite a few reviews about ChatGPT, a year later (after 2 major updates) it still serves as a benchmark. Sam Altman, at the head of this visionary company, has been an essential driver of these developments.
His departure therefore raises at a minimum questions about the future direction of OpenAI. Especially since the vote of the board that rejected it has certainly been prepared for a long time, which reveals turbulence in its governance, until the involvement of the main shareholders of which Microsoft. OpenAI’s other investors, Khosla Ventures, Sequoia and Thrive Capital, have reportedly expressed their astonishment at Altman’s sudden exit. The decision was unexpected, even among the people most closely connected with the company. Who is in charge?
Note that the governance structure of OpenAI is special. It involves a for-profit subsidiary that is fully controlled by a non-profit organization, since recall that OpenAI was created to promote open source AI. The non-profit organization has the power to determine when the company has achieved its general artificial intelligence (AGI) goal and can influence business conditions, including relations with major investors like Microsoft. In addition, Altman had no stake in OpenAI and could be fired at any time. He was certainly the first surprised, for not having protected himself against this risk.
According to the OpenAI Board of Directors, concerns about the transparency of its communications have been raised. These problems hindered the board’s ability to fully exercise its responsibilities, leading to a loss of confidence in its ability to lead the company. So there is a (big) “pan” in a cupboard, but we don’t yet know what it is, and how it will impact our projects around OpenAI technologies.
Even if we can reasonably think that his innovation capacity is not impaired by this departure, we can still wonder about his entrepreneurial capacity which is often the weakest link in the R & amp; D teams and yet famous labs. Sam Altman, before becoming the face of OpenAI, began his journey in the world of technology as co-founder of Loopt, a social geolocation platform acquired by Green Dot Corporation. His remarkable rise led him to the presidency of Y Combinator, a renowned startup accelerator, where he shaped the future of many start-up companies.
So we could think that he played his part in the launch of OpenAI and that now it is another skill that is needed to continue the road, a bit like Eric Schmidt arriving in 2011 at Google, after an exceptional career in the industry, at the request of the two founders. But we can also think that OpenAI has just lost its entrepreneurial engine. The announcement of the next CEO of OpenAI will therefore be closely scrutinized to see if it will be a captain of the industry, a renowned lawyer to manage the multiple open copyright disputes, or a technological player to control the growth of the platform and its security.
In this regard, OpenAI also blocked the creation of new paid accounts this week to deal with the success of its update, which also revealed possibilities to output data from the new custom ChatGPT… simply by asking nicely!
Because at OpenAI, Sam Altman was not just a CEO, when you read the interviews of him and his team in the press, where he has been on the cover many times, like that of Time this summer. He was the visionary who pushed the limits of artificial intelligence towards the general public and the “horizontal” uses of LLMs, but also concretized the ethical and practical questions about the use of these technologies. Awareness has risen to the oval office of the White House which is undertaking work on the governance of AI at the scale of the United States.
He has therefore clearly been a catalyst in the promotion of ethical and responsible AI, while advancing the state of the art of LLMs, making his vision that of OpenAI who will now have to communicate to reposition his vision, in a separate way from that of his founder. Apple has licensed Steve Jobs in 1985 and had to realign his product vision, to go down to hell… and to recall its creator to embody its vision in 1997. The history of Tech is filled with troubles, some of which are saving.
For GreenSI, the hypothesis of strategic non-alignment between Sam Altman and the board is close to the delicate balance between commercial imperatives and ethical commitments in the world of AI. OpenAI, as a non-profit entity with a commercial subsidiary, is at the intersection of these two worlds. Managing this duality is complex and transparency is a crucial aspect of it. Let’s remember the first departure of OpenAI, another co-financer, Elon Musk…
Beyond a certain uncertainty about OpenAI, not to say a certain uncertainty, it is still early to imagine, or even notice the impact of this departure on the new generative AI landscape. For GreenSI, as in a good game of chess, it will depend on the “next move”. We could, for example, fabulate about Elon Musk’s reaction with X.ai , Sam Altman’s next post, or Microsoft’s reaction that might want Copilot to become the next ChatGPT, even if it means killing ChatGPT… or why not Eric Schmidt also founder with Xavier Niel, call Sam in France 😉
The ability to retain OpenAI employees, in a very tight skills market, will also be a factor in redistributing the cards in the industry, as can be seen regularly between GAFAs who each seek “strategic catches” in the product managers. Such a significant change can also have an impact on morale within OpenAI, because governance gaps in companies are rarely limited to the hushed seats of the board.
Finally, the impact can also slow down high-growth investments if a period of uncertainty sets in.
So all the actors will be inclined to quickly create the “Post-Altman era”, because if investor confidence drops, not only in OpenAI, but potentially in the entire AI sector, it will not be good for anyone.