The Existential Question Behind the Musk-OpenAI Drama

It’s not just another celebrity drama. Or tech bros behaving badly. Not just a case of simmering resentments or personalities clashing. No, the dust-up between Elon Musk and OpenAI chairman Sam Altman is much more than that. It is, after all, based on an existential debate: Will artificial intelligence (AI) bring about the demise of all humanity?

That is the burning question at the center of the Elon Musk-OpenAI drama that recently culminated in a hostile takeover offer of $97.4 billion for the maker of ChatGPT.

But before scoffing at the notion that technology could eradicate humanity, consider the background. It might motivate Musk to publicly wage war against OpenAI and Altman as the business world continues to watch the battle play out against the backdrop of geopolitical and domestic tensions. It is this mindset that is propelling Musk to fight OpenAI today, according to his February 2024 lawsuit, which was withdrawn and then refiled in August of the same year.

Aron Solomon, chief strategy officer at Amplify, told PYMNTS, “We should actually care — very much — about the Elon Musk vs. OpenAI drama because it’s not just another billionaire feud. … This made-for-TV drama matters because AI is shaping everything — our jobs, our news, our social interactions, and even national security. If it’s controlled by a few corporations chasing a profit, that’s a problem.”

It all started, according to the 2024 suit, in 2012 when Musk was giving Demis Hassabis, the co-founder and CEO of DeepMind, an AI research lab, a tour of SpaceX’s Hawthorne, California, facility. They discussed the greatest threats facing society.

Hassabis told Musk that AI advancements are one such major threat. Computing circles have long theorized about an event called the “singularity,” where technology substantially exceeds human intelligence, with uncertain outcomes for humanity.

“Following this conversation with Mr. Hassabis, Mr. Musk became increasingly concerned about the potential of AI to become super-intelligent, surpass human intelligence, and threaten humanity,” according to the complaint.

Musk began talking about AI’s existential threat with his friends and colleagues, including Google co-founder Larry Page. But Musk was “shocked” that Page did not share his concerns, and allegedly said that AI systems replacing humans would merely be the “next stage of evolution.”

Musk discovered that Google was planning to acquire DeepMind, which was one of the most advanced AI companies in the industry. He became “deeply concerned that DeepMind’s AI technology would be in the hands of someone who viewed it and its power so cavalierly.”

Musk and Luke Nosek, PayPal co-founder, tried to buy DeepMind, believing that “the future of AI should not be controlled by Larry (Page).” They failed and Google acquired DeepMind in 2014.

The following year, Musk met Sam Altman, then president of startup incubator Y Combinator and a tech investor. Musk said he believed at the time that he found someone who shared his concerns. Altman would pitch the creation of a nonprofit AI lab filled with top researchers, one that would share its research with the world for free.

It was meant to be a counterweight to Google. The thinking was that giving the public access to state-of-the-art AI models would compete with — and defuse — Google’s attempts to develop superintelligence. Altman drafted a proposal for this nonprofit, which would own the technology and use it for the “good of the world,” according to the complaint.

Musk later came up with the name OpenAI, to reflect the lab’s open-source mission. It was 2015.

Musk, who pumped about $50 million in OpenAI to get it started, said he used his stature and wealth to poach respected research scientist Ilya Sutskever from Google. Musk and Hassabis were in a bidding war for him; but Sutskever would join OpenAI as chief scientist and co-founder.

Musk instructed Altman and OpenAI co-founder Greg Brockman to “get the best people in the world” and offered essentially a blank check: “Whatever it takes.” Musk added that DeepMind was “causing me extreme mental stress. If they win, it will be really bad news with their ‘one mind to rule the world’ philosophy.”

OpenAI started developing large language models. GPT-1, GPT-2 and GPT-3 were open source. But then they realized they needed more money to continue their work. “We all understood we were going to need a lot more capital to succeed at our mission — billions of dollars per year,” according to an OpenAI blog post.

It was at this point that Musk’s and OpenAI’s retelling of the story diverges.

OpenAI, which is seeking to dismiss all of Musk’s claims, said Musk didn’t come through with the $1 billion he promised to give OpenAI. OpenAI went to venture capitalists to raise funds, but found it a hard sell as a nonprofit, according to Brockman, who recounted what happened at a SXSW speech.

OpenAI said that Musk agreed it had to become a for-profit entity to continue its mission, but he also wanted control of OpenAI: majority equity, initial board control and to be named CEO. Musk also withheld funding while talks went on, so another investor stepped up, according to OpenAI. (Musk would exit in 2018.)

In 2019, OpenAI restructured to become a capped-profit company overseen by a nonprofit parent. In 2020, Microsoft licensed OpenAI’s models and let it use its cloud services to train AI models. Microsoft would reportedly invest a total of $14 billion into OpenAI, and own 49% of its capped-profit arm.

This development would infuriate Musk. OpenAI “has become a closed source, maximum-profit company effectively controlled by Microsoft. Not what I intended at all,” Musk said in a post on X.

In March 2023, OpenAI developed GPT-4, which showed initial inklings of AI’s holy grail, artificial general intelligence (AGI). AGI is achieved when machines can do a broad array of tasks, just like humans. GPT-4 tested in the 90th percentile for the uniform bar exam, among other milestones. But GPT-4 was closed and proprietary.

As OpenAI was stepping up its commercial activities, its board fired Altman in November 2023, saying they lost confidence in him because he had not been “consistently candid.” But under pressure from Microsoft and other investors, and employees who threatened to quit, Altman was reinstated less than a week later. The independent board members left, replaced mostly by corporate executives.

Meanwhile, Musk started his own AI startup, xAI, in March 2023 to compete with OpenAI. A year later, he would sue OpenAI for allegedly breaching its fiduciary duties. Musk said he invested in OpenAI under false pretenses — he was interested in the startup as a nonprofit, not a for-profit entity controlled by Microsoft, the most valuable company in the world at the time.

Musk offered to withdraw the lawsuit if OpenAI agreed to stay a nonprofit. The lawsuit, refiled last August, is ongoing.

In late December 2024, OpenAI said it would transition its capped-profit subsidiary into a for-profit public benefit corporation. This structure would require OpenAI to balance shareholder and stakeholder interests with the public benefit when making decisions. OpenAI said it is making the change to raise even more capital. The nonprofit parent would remain.

A month later, Altman would stand with President Trump, SoftBank CEO Masayoshi Son and Oracle Chairman Larry Ellison to announce to $100 billion to $500 billion project called Stargate to build AI infrastructure.

But the simmering fight that began seven years ago would erupt once again. Last week, Musk announced he wanted to acquire OpenAI for $97.4 billion. He offered to withdraw his bid if OpenAI didn’t convert to for-profit status.

Altman rejected the offer, instead cheekily offering to buy Twitter for “$9.74 billion,” according to a post on X. The OpenAI board would reject it officially later. OpenAI is also considering giving its nonprofit side special voting rights to further halt a hostile takeover by Musk, according to FT.com.

As the saga continues, Meta’s Chief AI Scientist Yann LeCun might have the answer to Musk’s fears about AI: Meta’s open-source flagship model Llama, which is free and performs at par with OpenAI’s and Google’s models. The public can build off Llama to offset any harmful superintelligence.

“And the open source counterweight (to OpenAI) is actually … [drum rolls] … Meta !!!” LeCun said in a post on X.

The post The Existential Question Behind the Musk-OpenAI Drama appeared first on PYMNTS.com.