Mustafa Suleyman recalls Larry Page's vision for interstellar probes

A meeting at SpaceX

Microsoft AI CEO Mustafa Suleyman recently shared a memory from a 2015 meeting held at the SpaceX headquarters. This gathering served as the very first meeting for the DeepMind AI Ethics and Safety Board. The meeting was a result of specific conditions set during the sale of DeepMind to Google in 2014.

Suleyman and his co-founders, Demis Hassabis and Shane Legg, required that Google establish an ethics committee with independent oversight. The board structure required mutual agreement between the two companies involved in the acquisition. This resulted in a diverse group of leaders overseeing the safety of the technology.

The board was composed of Google executives including Larry Page, Sergey Brin, and David Drummond. On the DeepMind side, Suleyman, Hassabis, and Legg were represented. The group jointly selected Elon Musk to serve as one of the three independent members for the board.

The lunch conversation

Following the morning session of the ethics board, the group sat down for lunch in the SpaceX canteen. Suleyman recalled sitting next to Larry Page during this meal. He noted that Page seemed suspicious of him during the interaction, even jokingly referring to Suleyman as one of those "speciesists."

During this lunch, Page shared a specific and ambitious vision for the future of human civilization. He spoke about the possibility of exploring distant galaxies by the end of the century. This vision was not about immediate commercial goals but rather a long-term technological trajectory for humanity.

Page specifically mentioned his desire to explore the galaxy at the head of a von Neumann probe. Suleyman recounted the conversation as a moment of deep, futuristic thinking. The exchange highlighted the types of goals that the founders of the modern internet were contemplating at that time.

Understanding von Neumann probes

A von Neumann probe is a theoretical type of robotic spacecraft designed for interstellar travel. These probes are intended to be self-propelling and capable of navigating across vast distances of light-years. They are designed to function autonomously without constant human intervention.

The core concept of these probes involves their ability to harvest energy and resources directly from space. By utilizing materials found in their environment, they could theoretically replicate themselves. This allows a single probe to create more probes, leading to exponential expansion across the galaxy.

While these probes are currently theoretical, they represent a significant milestone in space exploration science. They would allow human civilization to reach distant stars without needing to transport massive amounts of supplies from Earth. This technology would require advanced robotics and autonomous decision-making systems.

Elon Musk's endorsement

Elon Musk reacted to Suleyman's retelling of the 2015 conversation with strong support for the vision. He expressed that human civilization would be winning wondrously if these probes were being used by the year 2100. Musk has a long history of making bold predictions about the future of transportation and space.

Musk specifically noted that Larry Page might be right in his assessment of the timeline. He pointed to Page's track record of correctly predicting many major technological shifts over the decades. This endorsement reinforces the idea that the leaders of Google and SpaceX share a similar long-term outlook.

The conversation underscores the connection between AI safety and the ultimate goals of the tech industry. While the board was meant to discuss ethics, the personal visions of the members revealed a shared interest in deep-space exploration. It suggests that the technology being developed today is seen as a tool for much larger cosmic goals.

The legacy of the meeting

Suleyman noted that he remains unsure if these early discussions at SpaceX influenced the founding of OpenAI. However, the meeting serves as a significant historical marker for the intersection of AI ethics and space exploration. It shows how the architects of modern AI were thinking about the limits of human reach.

The involvement of Musk as an independent board member for DeepMind's ethics committee is a notable detail. It highlights the interconnectedness of the major players in the AI space. These relationships have shaped how safety protocols are developed for the most powerful models in existence.

As these companies continue to grow, the visions shared in 2015 remain relevant to their current trajectories. Whether through AI-driven robotics or autonomous spacecraft, the goal of expanding human influence into the galaxy remains a primary theme. The conversation at the SpaceX canteen serves as a snapshot of that ambition.