Five days of vertigo, if the events we have seen these days were an episode of Succession or a movie like Wall Street, it would seem to us that the scriptwriter has gone two towns… it would seem somewhat implausible, but reality often surpasses fiction. And in this case, the infighting at a San Francisco company with fewer than 400 workers threatens to shake the global balance of A.I.
It is a dramatic story, typical of a work of fiction. A superstar CEO betrayed by his Chief Scientific Officer at the height of his career. Some powerful allies and all the workers who fight for his return without achieving it. The unknown of what will happen to a company that, out of nowhere, has managed to reach 100 million users, 13 billion dollars from Microsoft and for some the most important threat to democracy and humanity in a world where armed conflicts bloodthirsty are an everyday occurrence. And a happy ending in which the protagonists win. The perfect Holywood script!
Many of you have probably not followed the saga and some of you have never used ChatGPT or DALL-E, but it is also a story of innovation, business policy and even geopolitics that shows like no other that it is not technology but the decisions of human groups and their power struggles, which build history.
Last Friday, November 17, OpenAI’s board fired its charismatic CEO, Sam Altman, by surprise. The explanations were few, brief and opaque, mentioning that Sam Altman had not been completely honest with them…. in what? A mystery.
From here on, a frenetic set of events takes place. Among them, the pressure from its main investor, Satya Nadella, CEO of Microsoft, for the decision to be revoked. Altman’s new negotiations with the council. Council resignation proposals. The resignation of his second, Greg Brockman. The hiring of both by Microsoft. And finally a letter signed by practically the entire staff asking for Altman’s return.
Some of these events have had dramatic overtones, such as the unexpected change of position of the instigator of the coup – Llya Sutskever, the Scientific Director of OpenAI – who wrote in Greg Brockman. He also signed a letter along with the one who had been appointed interim CEO, Mira Murati (OpenAI’s Chief Technology Officer) and 550 of the 770 OpenAI workers asking for the return of both or threatening to leave the company. Finally, the board resigns, Sam Altman returns to the company and a new board is elected with prestigious names such as Larry Summers, the most charismatic President of Harvard with enormous prestige.
OpenAI’s atypical structure is what has made it possible, but not its cause.
In fact, OpenAI was born as a non-profit with a structure focused on doing things in a different way, with an ethical orientation, with the aim of taking the new step towards artificial intelligence that truly benefits humanity. Your “board” reflects that orientation. As time went by and as the computational needs increased and had to be paid for, a second structure appeared – this time, for profit – but where the OpenAI board still has a majority and where the profits of the companies that invest are limited.
This atypical structure is what allows a very small number of people who joined the OpenAI board a few years ago to still control the organization and be able to oust its CEO.
However, the causes should not be sought in the structure of OpenAI but in the confrontation of ideologies.
Much of OpenAI’s advice is close to the “effective altruism” movement (https://www.effectivealtruism.org/), a movement that seeks a type of altruism that has the maximum possible impact on society. One of their fundamental concerns is the impact of technologies, such as A.I. generative, which could, in his opinion, endanger the future of humanity.
Obviously, the rapid development of generative Artificial Intelligence and, especially, the fact that it was launched on the market so that users could adopt it in a very gradual way, was not to the liking of these currents of opinion and created confrontation. Globalization through the Internet polarized these positions to unsuspected extremes throughout the world, but especially in OpenAI. It is in this polarization where lies the origin of the conflict in which the peculiar structure of OpenAI provided the opportunity for an action of “effective altruism” in favor of humanity, ousting Sam Altman and slowing down development at OpenAI with the idea that This slowdown would help reduce the risk of endangering the planet.
In short, OpenAI had to be stopped to save the world.
This is not a situation that anyone would have wanted, but as in all things, there are winners and losers.
There are two clear winners, Satya Nadella, possibly the most brilliant executive, along with Elon Musk, in the current universe. And Sam Altman who has advice tailored to his ambition and his roadmap.
Whatever the outcome of the OpenAI, Sam and Greg saga, Microsoft won. If they had ended up at Microsoft, Microsoft would win, if – as has been the case – they would have managed to return to OpenAI and take control, Microsoft’s position would be strengthened, if they had created a new company, Microsoft would win, but less. It is therefore admirable the way in which Satya Nadella has handled a highly complex situation.
The second winner is Sam Altman and of course his team with Greg, … who now have a board that lives up to their ambition and is aligned with their roadmap alongside a Microsoft that is more aligned than ever.
For OpenAI and generative artificial intelligence there are two futures.
The first is already underway, it is that of innovation. The development of generative Artificial Intelligence and its first steps, such as ChatGPT, DALL-E or Midjourney, have opened the doors of an explosion of innovation that will flood us with proposals at least for some time.
The first, those who are already here, are the Copilots. Assistants that help us write, program, develop PowerPoints, client proposals and Excels. The leader in all this is an unexpected leader, Microsoft, and it is so on its own merit.
Then will come the specialized Copilots and the company Copilots. Many of them are already in development in leading companies such as Bain, PwC, Amazon, etc… we will see them as legal assistants, in human resources, in consulting and, above all, in software development, but also in industry.
Some developments will take a little longer, but will have a significant impact. Such is the case of automating customer service with chatbots, this time, intelligent and capable of starting conversations. We will see them in banking, in telephone companies and also in governments and cities. They will have a significant impact on employment, call-centers will be a thing of the past.
The second future, also underway, corresponds to new developments in generative artificial intelligence. Increasing your reasoning and calculating abilities seems within reach.
But even more within reach is the incremental innovation that will allow us to provide them with image, voice and personality, turning them into true agents with whom we can interact naturally. Or, provide them with mechanisms with which they can act and fill out forms, answer emails or attend meetings for us. This will lead us to unexpected developments in fields such as health, education and services that require or manage information, transforming a good part of our lives and our society.
The latest events point to an OpenAI that, together with Microsoft, will go faster and have more influence. Those who wanted to stop OpenAI to stop history could not have done worse.
