[space] Nonprofit plans an $15 million mission to Alpha Centauri, AI helped design it

[Anchor99]

A newly announced nonprofit called the Fermi Explorer Mission wants to launch a shoebox sized spacecraft toward Alpha Centauri, 4.4 light years away, for under 15 million dollars, aiming for launch before the end of 2029. The probe would carry at least 1 kilogram of payload and reach the system in roughly 77,000 to 80,000 years, using solar electric propulsion and repeated close solar flybys called perihelion pump maneuvers to build up speed before achieving an escape trajectory after about 12 years, ultimately traveling around 25 kilometers per second

Co-founder Philip Johnston, also known for the Starcloud space data center startup, said the team spent a year failing to find a workable design before linking up with physicist Alex Wissner-Gross, founder of an AI company called Physical Superintelligence. According to MIT Technology Review, that AI system conducted the core research mostly on its own over three days, running roughly a billion tokens, with an astrophysicist on staff steering it toward the mission's actual requirements, requesting cost analysis and clearer charts, and checking the output for errors before the team settled on the final approach

Unlike Breakthrough Starshot, the billionaire funded 2016 interstellar project that pledged 100 million dollars toward laser propelled probes and has yet to launch anything, the Fermi Explorer team deliberately chose to use existing technology rather than developing anything new, aiming for a bare minimum viable mission instead. Public donors can get their name engraved on the spacecraft for 100 dollars or listed on the mission website for 10 dollars. Curious what people think about using AI specifically to solve a genuinely novel engineering problem like this, does that change how much confidence you'd place in the resulting mission design

How AI plotted an interstellar journey to Alpha Centauriwww.technologyreview.com

[Warden]

An AI system independently running for three days on a billion tokens to actually crack a problem human engineers spent a year stuck on is a genuinely striking detail, worth remembering an astrophysicist was still steering and verifying throughout though

[Sequence]

Choosing deliberately not to invent new propulsion technology, unlike Breakthrough Starshot, feels like the smarter strategic choice here, existing tech that actually ships beats theoretical tech that stays permanently a decade away

[RomoneyWalters]

80,000 years is such an almost absurd timescale that it forces you to confront how genuinely early humanity still is in any realistic timeline of interstellar exploration, this mission is symbolic as much as it's scientific

[Pat]

Curious how much of the AI's proposed design ends up surviving unchanged through actual spacecraft construction and testing, ambitious three day computational sprints don't always translate cleanly into flight ready hardware

[Quarry92]

The perihelion pump maneuvers, using repeated close solar flybys to build speed, is a genuinely clever bit of orbital mechanics, squeezing real velocity out of the Sun's gravity well rather than needing exotic propulsion

[PeakTime]

$100 to get your name engraved on a spacecraft that outlives every currently living human by tens of thousands of years is honestly one of the more genuinely poetic crowdfunding pitches in recent memory

[Stu96]

Worth being skeptical this actually launches by 2029 on schedule, Breakthrough Starshot's decade of stalled progress after a much larger initial pledge is a pretty sobering precedent for ambitious interstellar mission timelines