Elon Musk's 2017 Bid to Lure Sam Altman to Tesla's AI Ambitions
*Newly revealed messages show Musk's team pitching OpenAI's co-founder to lead a rival lab, highlighting early tensions in the AI race.*
Elon Musk's team at Tesla floated the idea of recruiting Sam Altman, the co-founder and CEO of OpenAI, to head up a new AI lab in 2017. This move came as Musk sought greater control over OpenAI, the nonprofit AI organization he helped start but later clashed with over its direction.
OpenAI launched in 2015 with Musk as a key backer, aiming to advance artificial general intelligence in a way that benefited humanity. By 2017, however, Musk grew concerned about the pace of AI development at rival labs like Google's DeepMind. He had already expressed frustrations with OpenAI's structure, which prioritized open research over proprietary applications. Tesla, meanwhile, was ramping up its own AI efforts for autonomous driving, relying on neural networks and computer vision.
The messages, exchanged between Shivon Zilis—a Neuralink executive and early OpenAI advisor—and Tesla executives, outline plans for this rival AI initiative. Zilis, who worked closely with Musk on AI projects, discussed the possibility of Altman leading the lab. As an alternative, they considered Demis Hassabis, the founder of DeepMind, known for breakthroughs in protein folding and game-playing AI. These discussions happened amid Musk's push to shift OpenAI toward for-profit models, a pivot that eventually led to his departure from the board in 2018.
The Messages and the Pitch
Zilis's communications reveal a strategic scramble. In one exchange, she highlighted Altman's potential to bridge Tesla's hardware expertise with cutting-edge AI research. Altman, then in his early days as OpenAI's leader, had a track record in startups like Loopt and Y Combinator, making him appealing for Tesla's ambitions in self-driving tech. The pitch emphasized creating a lab that could compete directly with OpenAI, focusing on applied AI rather than pure research.
Tesla executives saw this as a way to accelerate Autopilot development. At the time, Tesla's AI team was small, handling vast amounts of driving data but lacking the theoretical depth of academic labs. Recruiting Altman would have brought not just expertise but also talent pipelines from OpenAI's network. Hassabis emerged as a backup; his work at DeepMind had demonstrated scalable reinforcement learning, which could apply to real-world robotics and vehicles.
The plan didn't materialize. Altman stayed at OpenAI, which transitioned to a capped-profit model in 2019 to attract investment. Musk, frustrated, sued OpenAI in 2024, alleging it abandoned its nonprofit roots. These 2017 messages, surfacing now during the ongoing trial, paint Musk's recruitment effort as a calculated countermove.
Broader AI Landscape in 2017
This episode underscores the talent wars brewing in AI even then. Companies like Google, Facebook, and Microsoft were pouring billions into research, luring top minds with equity and resources. Musk's interest in Altman reflects his pattern of centralizing control—Musk had co-founded OpenAI partly to counter Big Tech's dominance but chafed when it didn't align with his vision for Tesla and SpaceX.
Zilis's role adds context. As a bridge between Musk's ventures, she advised on AI ethics and talent. Her messages suggest Tesla viewed the lab as an extension of its Dojo supercomputer project, later unveiled for training AI on driving data. Without Altman or Hassabis, Tesla built its team internally, hiring from academia and competitors.
No public response from Altman or Hassabis appears in the messages. OpenAI has since grown into a powerhouse, valued at over $80 billion, while Tesla's AI focuses on practical deployments like Full Self-Driving software.
Reactions from the Players
Musk has not commented directly on these specific messages in recent statements. During the trial, his lawyers argue OpenAI misled him about its nonprofit commitment. OpenAI counters that Musk pushed for control, including a merger with Tesla, which the board rejected to maintain independence.
Zilis, now deeply involved in Neuralink's brain-machine interfaces, has kept a low profile on the matter. Tesla's current AI head, Andrej Karpathy, who joined in 2017 around the same time, has praised the company's scrappy approach but acknowledged the challenges of scaling without top external hires.
Critics in the AI community point to this as early evidence of Musk's combative style. Some ethicists worry such poaching tactics fragment efforts on safe AI development, while proponents see it as healthy competition driving innovation.
Why This Matters
These revelations matter because they expose the personal and strategic fault lines that shaped today's AI giants. Musk's 2017 gambit failed, but it foreshadowed the schisms: OpenAI's pivot to commercial models under Altman fueled its rapid growth, while Tesla doubled down on proprietary AI for hardware integration. For engineers and founders, this is a reminder that talent isn't just skill—it's about alignment with a founder's vision.
In the current landscape, where AI labs battle for scarce experts, such recruitment plays can make or break trajectories. Musk's effort didn't sway Altman, but it accelerated Tesla's internal push, contributing to tools like the Optimus robot. OpenAI, free from Tesla's constraints, pursued broad AGI research, leading to models like GPT-4. The trial may unearth more, but this glimpse shows how interpersonal dynamics, not just tech, dictate AI's path.
Ultimately, the messages affirm that control in AI often hinges on who leads the lab. Musk lost that round, but his influence persists across ventures.
---
No comments yet