To enjoy additional benefits
CONNECT WITH US
November 26, 2023 12:15 am | Updated 01:39 am IST
The story so far: OpenAI, the company that is at the cutting edge of the AI revolution, almost ceased to exist over the last weekend. Sam Altman, the founder and CEO of the company that runs ChatGPT, was ousted by its board of directors on Friday, November 17 only to be reinstated on Tuesday, November 21, and the board unceremoniously disposed of. The intervening few days saw two interim CEOs, a bristling Satya Nadella, a gleeful Elon Musk, and over 700 employees in open revolt. The cause of this chaos can be traced to a contradiction that exists at the heart of OpenAI.
OpenAI was founded in 2015 as a non-profit organisation by a group that included the then 30-year-old Sam Altman, Elon Musk, and Infosys, among others, with the vision that Artificial Intelligence research must be kept open, safe and available to everyone. A note dated December 11, 2015 on OpenAI’s website reads: “Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return.” Incidentally, this note was penned jointly by Greg Brockman, another co-founder who was also fired from his post of Chairman of the board on Friday, and Illya Sutskever, OpenAI’s chief scientist and board member, who was reportedly a main organiser of the boardroom putsch. OpenAI may have spurned financial returns but AI research is a costly affair. The Generative Pre-Trained Transformers or GPT models that sit at the heart of many AI tools need to be trained on vast amounts of data, which require massive amounts of ‘compute’ or computing power. The compute at the scale required by large AI models can be bought from server farms operated by corporations such as Amazon, Microsoft and Google.
In 2018, OpenAI’s biggest source of funds dried up when Mr. Musk left the board citing a conflict of interest with Tesla. A year later, OpenAI under Mr. Altman’s supervision started a for-profit wing that would go on to build monetisable consumer AI technology such as ChatGPT and Dall-E. Mr. Brockman and Mr. Sutskever put out another note at that time saying that such a venture was necessary so that the company could “invest billions of dollars in upcoming years into large-scale cloud compute, attracting and retaining talented people, and building AI supercomputers”. The for-profit venture essentially reported to the non-profit organisation led by a board that is hypersensitive about the commercial exploitation of AI. This tussle between two opposing principles in the same organisations came to a head with the release of ChatGPT for free in November 2022, and the public got a taste of what AI is capable of. The flood of users that followed brought with it investors as well.
Microsoft was with OpenAI before it had made a name for itself. Between 2019 and 2022, it is reported to have pumped in up to $3 billion in Open AI — an investment that appeared prescient once ChatGPT appeared on the scene and OpenAI’s valuation shot up to $29 billion. In January 2023, Microsoft added $10 billion to the kitty, mostly as much-needed computing power from its Azure cloud computing platform. CEO Satya Nadella was lauded as a visionary for the early investment and Microsoft went on to integrate ChatGPT into its Bing search engine — a move one step ahead of search giant Google, which was once seen as the leader in AI research. Given its level of involvement in OpenAI, the board’s coup attempt on Friday blindsided Microsoft. While Mr. Nadella stated that he will continue to work with OpenAI, he made it abundantly clear that he wanted Mr. Altman back at the helm. When an initial round of negotiation between Mr. Altman and the board the day after he was fired failed, Mr. Nadella announced that he would hire Mr. Altman and Mr. Brockman to run an AI research centre within Microsoft, along with any OpenAI employee that jumped ship. This is exactly what 702 staffers, including CTO Mira Murati who was initially appointed as interim CEO, threatened to do if Mr. Altman and Mr. Brockman were not brought back and the board did not quit. The employees may have been driven by loyalty, as well as the knowledge that their stock options may drop significantly without Mr. Altman’s commercial drive. With OpenAI’s creations sitting exclusively on Azure platforms and all of its employees on board, Microsoft would have essentially “acquihired” the most important AI company of the moment. However, that would also mean it would have to take direct responsibility for any AI goof-ups — which government regulators are keenly watching.
After its initial CEO pick Ms. Murati sided with Mr. Altman and pushed for his return, the OpenAI board roped in Emmet Shear, a co-founder of the Twitch streaming platform, as interim CEO. However, internal divisions seem to have developed in the board, with Mr. Sutskever repenting his role in the fiasco and signing the employees’ letter demanding Mr. Altman’s return. The pressure from employees and investors and its own divisions seemed to finally have an effect, as the board announced the return of Mr. Altman as CEO on Tuesday and its own reconstitution, with all the coup instigators except one — Quora CEO Michael D’Angelo — having resigned. Mr. Sutskever has been dropped from the board but reportedly remains in his employee position.
While it may seem like a place led solely by money and technology, Silicon Valley also has its philosophies, a key one being ‘effective altruism’. Simply put, effective altruism looks at ways in which any intervention, monetary or technical, can be most effective. The practitioners of this philosophy among the Silicon Valley elite have an obsession with the possible negative impacts of AI and reducing the risk associated with it. OpenAI’s charter itself speaks of safely building an AGI, or Artificial General Intelligence, an AI capable of reasoning like humans unlike the generative AIs that we have that only create based on what it has ‘learned’. Effective Altruists tend to push back against the ‘techno optimists’ and ‘accelerationists’, who believe that the benefits of technology outweigh the bad and that all technological developments need to be accelerated as it is the way forward for mankind. The effective altruism proponents on OpenAI’s board seem to have been spooked by the rapid commercialisation of the company and feared that it was deviating from its original purpose, playing into the hands of accelerationists. They may have been trying to recapture the narrative but bungled up the effort.
BACK TO TOP
Comments have to be in English, and in full sentences. They cannot be abusive or personal. Please abide by our community guidelines for posting your comments.
We have migrated to a new commenting platform. If you are already a registered user of The Hindu and logged in, you may continue to engage with our articles. If you do not have an account please register and login to post comments. Users can access their older comments by logging into their accounts on Vuukle.