AI-designed solar-bypass trajectory plots 80,000-year voyage to Alpha Centauri
The Fermi Explorer Mission plans a late 2029 spacecraft launch using a novel Sun-grazing maneuver discovered by physical-AI startup "Get Physics Done."
Original cover art, generated for this story. THE VISSION does not republish third-party press imagery.
- The Fermi Explorer Mission announced a late 2029 spacecraft launch to Alpha Centauri, utilizing a highly unconventional trajectory plotted by artificial intelligence.
- The trajectory, discovered by physical-AI startup 'Get Physics Done,' utilizes a Sun-grazing pass closer than Mercury to gain a high-speed energy boost.
- Developed by Physical Superintelligence (PSI), a lab backed by Bill Gates's Breakthrough Energy, the AI spent three days analyzing physics equations using a billion tokens.
The Fermi Explorer Mission, a private nonprofit organization, has announced plans to launch an interstellar spacecraft to the Alpha Centauri star system by late 2029, utilizing a highly unconventional, energy-maximizing trajectory discovered entirely by artificial intelligence, MIT Technology Review reported on September 1. The ambitious $15 million privately funded mission will carry scientific instruments and a copy of the historic Voyager Golden Record.
The trajectory was designed using 'Get Physics Done,' an open-source AI system developed by Physical Superintelligence (PSI), a newly launched physical-AI research lab that recently secured $58 million in funding led by Bill Gates’s Breakthrough Energy. To calculate the path, the model spent three days running physics simulations and consuming approximately one billion tokens under the steering of human astrophysicists.
The AI model suggested a high-risk, high-reward maneuver that human astrophysicists had not previously modeled: the spacecraft will swing closer to the Sun than the orbit of Mercury. By firing its engines at the point of closest solar approach, the craft will utilize four times the solar intensity and harness the Oberth effect — a physics principle where a rocket engine's high-speed thrust is significantly more efficient at high velocities. This Sun-grazing trajectory allows the spacecraft to remain lightweight while carrying small, low-mass solar panels.
The journey is estimated to take up to 80,000 years to reach the nearest star system, which is 4.4 light-years away. Philip Johnston, president of the Fermi Explorer Mission, acknowledged that faster spacecraft built in the future will likely overtake the probe, but argued that the priority is to execute an actual launch rather than wait for theoretical high-speed propulsion that may never materialize.
Having an AI successfully discover a Sun-grazing solar-bypass trajectory that human orbital mechanics experts overlooked is a powerful proof of concept for 'physical AI' models trained on formal scientific laws. By demonstrating that LLMs fine-tuned on physics can discover valid, novel engineering solutions within a billion tokens of compute, Physical Superintelligence shows that AI's most profound role in the hard sciences is as a tool for extreme engineering optimization.