As a six-sided warrior in the tech world, Elon Musk simultaneously leads SpaceX, Tesla, Neuralink, and other companies.
If not for his recent lawsuit against OpenAI’s current CEO Sam Altman, many might not have known about his deep connection to OpenAI – he even named the organization.
The complex relationship didn’t fade with time. Recently, Musk released a batch of email exchanges with OpenAI executives. While these may not tell the complete story, these key communications help piece together a more complete picture of their relationship.
Before diving in, let’s introduce the key players in these emails:
- Sam Altman: Current CEO of OpenAI
- Elon Musk: OpenAI co-founder, CEO of SpaceX, Tesla, Neuralink, etc.
- Greg Brockman: Current President of OpenAI
- Ilya Sutskever: Former Chief Scientist at OpenAI
- Andrej Karpathy: OpenAI founding member, former Tesla AI Director
- Shivon Zilis: Former OpenAI advisor, joined Tesla in 2017, close to Musk
The Birth of OpenAI: A Summer Beginning
In May 2015, Silicon Valley was quietly but inevitably being transformed by the AI wave.
During a meaningful email exchange, Sam Altman and Elon Musk realized that AI development was unstoppable. Rather than letting Google dominate, they decided to forge their own path.
“We need a non-profit organization,” Altman proposed, “to attract top talent who truly care about AI’s direction.”
By summer 2015, Altman began designing OpenAI’s governance structure. He envisioned a five-person core team, with technology rights belonging to the foundation for global public benefit. Musk gladly accepted a “part-time partner” role.
Safety was positioned as OpenAI’s top priority. The core team would carefully decide which research to open-source and which to keep private.
One month before OpenAI’s official launch, Greg Brockman joined the discussion, advocating for OpenAI to enter the AI field as a neutral institution.
Three days before launch, the focus shifted to talent recruitment strategy. Musk emphasized that OpenAI’s foundation was its mission to “benefit humanity,” seeing this as key to attracting top talent. Altman quickly adjusted the organization’s messaging accordingly.
However, on launch day, Altman worried about competing with Google DeepMind’s superior compensation packages. He suggested raising salaries to remain competitive in the talent war.
Musk focused on recruiting key figure Ilya Sutskever, offering to participate in recruitment full-time to secure his talent. Ilya’s eventual joining delighted Musk.
On December 11th, Musk sent an excited email to OpenAI’s early elite team:
Congratulations on our brilliant start!
While our resources may be limited compared to some organizations you know, we have righteousness on our side, which is crucial. I’m optimistic about our chances. Our most important task is recruiting the best talent.
A company’s results reflect its employees’ capabilities. If we can attract the most talented people and maintain the right direction, OpenAI will surely succeed.
The Battle Between Business and Ideals
Entering 2016, the team began refining operational plans. Early full-time employees received 275,000 annual salaries plus 0.25% YC company shares. New employees could choose between 175,000 salary + $125,000 annual bonus or equivalent YC or SpaceX stock.
Despite seeming generous, these packages were below industry standards. Most core technical staff took pay cuts to join OpenAI. Even intern compensation was below market rates.
However, facing fierce competition from DeepMind, Musk expressed willingness to reassess compensation to attract global top talent.
During this period, OpenAI’s external communications largely followed Musk’s vision. When Brockman interviewed with WIRED, he needed to confirm key points with Musk to ensure consistency with his views.
In September 2016, Microsoft offered an attractive partnership: 60 million in computing resources for a 10 million investment from OpenAI. However, it required OpenAI to evaluate and optimize Microsoft products and endorse Azure.
Musk immediately opposed and vetoed this proposal. Eventually, through Altman’s mediation, OpenAI reached a $50 million agreement with Microsoft without mandatory conditions.
Internal Turbulence
By late summer 2017, important discussions about the organization’s future became a crucial turning point.
Brockman and Ilya had a weekend discussion sharing their concerns, which reached Musk through advisor Shivon Zilis, triggering a storm that would alter the organization’s fate.
They raised several points:
- No individual should control AGI if developed
- Questioned Musk’s time commitment to OpenAI
- Sought more support in hardware areas
- Requested expanded employee equity pool
This “secret meeting” content enraged Musk. “This is really annoying,” he wrote in his reply, “encourage them to start a company. I’ve had enough.”
In September, tensions escalated further. Ilya expressed concerns about Musk potentially becoming an AGI dictator and opposed using Tesla resources to acquire AI chip company Cerebras.
By early 2018, OpenAI faced funding challenges. When Altman suggested an ICO for fundraising, the security team raised alarms.
Karpathy shared concerning data showing Google’s dominance with 83 paper submissions at top deep learning conferences. Musk pointed out: “OpenAI is heading toward inevitable failure compared to Google.”
Musk and Karpathy advocated merging OpenAI into Tesla as the only way to compete with Google. However, OpenAI chose a different path, creating an innovative funding model: maintaining OpenAI Inc. as non-profit while establishing OpenAI LP as a profit-capped entity.
By March 2019, OpenAI’s new structure was complete, with investor returns capped at 100 times investment. Musk had already fully withdrawn, as noted in Altman’s press release: “Elon Musk left the OpenAI Nonprofit board in February 2018 and is not involved with OpenAI LP.”
The final disclosed email from March 11, 2019, showed Musk requesting Altman to clarify to Bloomberg that he had no financial relationship with OpenAI’s profit-making subsidiary. Altman simply replied “understood.”
This marked the end of an era and the beginning of another. The organization created to benefit humanity continues its mission, though under Altman’s leadership, the path has proven more complex than imagined.
These emails reveal a group of innovators with grand ambitions. Technological progress is never smooth, and the path to AGI remains long and challenging.