Open AI Drama. An interpretation.
Laid off on a Friday, snapped up by Microsoft on a Monday, then back at the helm of his company on a Tuesday evening: Sam Altman's life has recently gone through ups and downs that few humans will ever experience. Like the technologies themselves, the world of AI companies is going through an unusual growth crisis, unraveling the usual space-time continuum of start-ups. The classic dichotomy between the technical or entrepreneurial profiles of the founders, and the ultimately more public responsibilities induced by the societal impact of these technologies, is intertwined this time with another problem: the divergent objectives within the founding team, and the tensions between the non-profit and the commercial structures, as well as the very philosophies underpinning them.
From the outset of the OpenAI project, two philosophies have coexisted: the first is what the Americans call Effective Altruism. This goes beyond AI, as it can be found, for example, in the fight against poverty or the defense of animal welfare. Proponents of this theory, instead of donating to charitable causes, are themselves striving, within the corporate framework, to define objective variables to determine the societal impact of their actions. In the context of artificial intelligence, the board of directors of the OpenAI Foundation claimed to manage the issues inherent in AI, such as those of alignment, safety, and the advent of general-purpose AI (AGI), through rules internal to their company. Such an approach implied an adapted pace of business innovations in order to study their consequences. Until Friday, this Board saw itself as a Council of AI sages. Other OpenAI executives, just as idealistic about the applications of this technology in healthcare, engineering, etc., endorsed Effective Accelerationism: a more enthusiastic technological trend wishing to move faster with change. In both cases, we note the contradiction of striving for quasi-humanitarian aims, while developing technologies requiring colossal amounts of capital, provided in this case by venture capitalists and then by Microsoft. By acting as a think-tank and regulatory foundation, while Open AI had become a commercial enterprise over the past two years , the Company had strived to pionneer a model of corporate governance that has been flawed from the outset. Once it passed the billion-dollar mark in annual revenues, OpenAI become a company like any other, and must now abide by the same rules. This is not to say that the Foundation has no right to exist, but it must separately continue its existence. As for the opposition between the two approaches to technological change, altruistic or commercial, this must be discussed in public debate, with regulators, philosophers and economists, and not simply between players in the sector.
Sébastien Laye is a French American economist and entrepreneur