Machine Explorers

Will A.I. Replace Astronauts?

Launching autonomous systems into space may define the longevity of our technological civilization. Have others gone before us?

Avi Loeb
Point of Contact


Images: NASA — Challenger astronaut | Perseverance Rover

The Perseverance rover surveys Mars as a six-wheeled motorized robot, controlled by the Jet Propulsion Laboratory (JPL) in Pasadena. The level of control from Earth would have been toned down considerably if an astronaut was in charge of operating the associated instruments.

Will there be a time when artificial intelligence (AI) systems replace astronauts and make autonomous decisions based on their own machine-learning experience? This feat will establish a phase transition in our ability to promote humanity’s goals in space. Traversing interstellar distances with chemical rockets requires millions of years. In contrast to the fragile nature of living organisms, technological equipment would be resilient to the harsh interstellar conditions, which bring about bombardment by energetic cosmic-rays as well as irradiation by harmful UV and X-rays.

It takes light a long time to traverse interstellar distances, at least four years from our nearest star. As a result, communicating timely signals from Earth to an interstellar probe is impractical. The craft must possess its own intelligence and operate independently, following the general guidelines it received on Earth early in its life — similarly to the way our kids explore the world on their own after growing up.

There is tension between our hope to cross this technological threshold of interstellar travel by autonomous AI systems and the concern that our civilization will end its own life earlier than that as a result of self-inflicted wounds, like climate change, technological catastrophes or a non-conventional war.

Given the slow response of policy makers to the existential risks that stem from terrestrial technologies, our civilization might only have a few centuries left for it to accomplish interstellar AI presence. It is crucial that we develop the capacity to launch autonomous AI systems to space within the near future. Otherwise, all our eggs — which are currently kept in a single basket we call Earth — could be lost in a single local catastrophe.

Most people do not recognize the race between self-destruction and interstellar AI presence as an existential threat for creating an outpost in space that would outlast us.

But this race must characterize any technological civilization out there. The abundance of space relics from technological civilizations that predated us in the Milky Way galaxy, depends crucially on whether these civilizations crossed the same threshold or not.

Another important technological milestone is whether their spacecraft have the ability to reprint damaged parts with 3D printers or even replicate themselves, as biological systems do. A deep wound imprinted on my body last year is invisible by now. Here’s hoping that one day, we will develop a car that would heal itself after an accident. This healing ability would be very helpful for self-replicating machines in space, of the type envisioned by John von Neumann and others.

At the other end of the technological survival spectrum, the relics from civilizations which were buried under the rubbles of self-destruction before sending self-replicating AI systems to space would be much more difficult for us to trace.

The recent emergence of technology on Earth is very rapid, resembling an exponential instability. Once we started developing them, our technologies advanced fast on cosmic timescales. But the means for our destruction grow in parallel to the means for space exploration, and so it is unclear which of these trends will prevail.

Most technological civilizations may have perished before venturing into space. But it is sufficient to have merely one advanced civilization which managed to overcome the aforementioned threshold for interstellar AI presence, for it to shape our Galactic landscape. If it started a billion years ago, it could have populated by now all habitable planets in the entire Milky Way galaxy with self-replicating AI systems, including the neighborhood of the Earth in the Solar system.

To find out whether we live in such a reality, we should search for technological equipment floating in the Solar system that is not human made. This is precisely the objective of the recently announced Galileo Project. Finding what happened to others through telescopes would help us plan for our own long-term survival in the future.

Studying autonomous AI systems from extraterrestrials may inspire our ambitions in space as well as on Earth. It may provide the secret recipe for our longevity. This could be the key to humanity’s future in the Darwinian selection process for the survival of the fittest species within the vast and old cosmos that surrounds us.


Avi Loeb is the founding director of Harvard University’s — Black Hole Initiative, director of the Institute for Theory and Computation at the Harvard-Smithsonian Center for Astrophysics, and the former chair of the astronomy department at Harvard University (2011–2020). He chairs the advisory board for the Breakthrough Starshot project, and is a former member of the President’s Council of Advisors on Science and Technology and a former chair of the Board on Physics and Astronomy of the National Academies. He is the bestselling author of “Extraterrestrial: The First Sign of Intelligent Life Beyond Earth” and a co-author of the textbook “Life in the Cosmos.”



Avi Loeb
Point of Contact

Avi Loeb is the Baird Professor of Science and Institute director at Harvard University and the bestselling author of “Extraterrestrial” and "Interstellar".