Four astronauts flew around the moon last week. Reid Wiseman, Victor Glover, Christina Koch, and Jeremy Hansen completed a 10-day lunar flyby aboard Orion, the first crewed lunar mission since Apollo 17 in 1972. The thousands of engineers who built that capsule, who welded the SLS core stage, who ran the ground systems at Kennedy, who debugged software at 2 a.m. for years: they pulled it off. That is real. That deserves to be said plainly.
And then you look at the number: $4 billion per launch. Four times the original estimate. The Space Launch System costs $2.2 billion just for the rocket, which burns up in the ocean after one use. One flight. Gone. At that price, the $93 billion total program estimate through 2025 stops being an abstraction and starts being a policy failure.
The Engineering Case Against Keeping SLS
The honest tension here is that I want both things to be true. I want to celebrate what Artemis II accomplished and I want to argue that the vehicle that carried it is the wrong tool for what comes next. Those two positions are compatible. The Wright Flyer was a genuine achievement. Nobody proposed scaling it to carry passengers.
NASA Administrator Jared Isaacson said it directly in February: SLS is not the vehicle you use to build a moon base at cadence. He is right. The metric that matters for sustained lunar presence is not whether you can launch once; it is cost per kilogram to orbit and how many times per year you can fly. SLS fails both tests badly. SpaceX's Starship, still maturing, targets a fraction of the per-launch cost and is designed for full reusability. That is the engineering trajectory that leads somewhere.
Critics of commercial space will point out that Starship has not yet carried crew to the moon, and that is a fair point. Unproven systems should not replace proven ones overnight. But NASA already awarded SpaceX the Human Landing System contract, which means Starship is already inside the Artemis architecture. The question is not whether commercial systems belong in lunar exploration. They are already there. The question is why Congress keeps funding a $2.2 billion single-use rocket alongside them.
The Political Geometry Protecting a Bad Investment
Artemis suppliers operate in all 50 states. That is not an accident; it is a procurement strategy designed to make the program politically indestructible. Every senator has a constituent whose job depends on SLS continuing. Mike French of Space Policy Group described it accurately: key congressional members protect the program because the program was built to need protecting. That is not aerospace engineering. That is defense contracting logic applied to exploration.
NASA's FY2026 budget still requests $2 billion for SLS and $1.37 billion for Orion. The White House proposal asks for vague assessments of commercial alternatives for missions after 2028, which is not a plan. It is a delay with paperwork attached.
Here is what should happen: Congress should set a hard cutoff date for SLS after Artemis III, fund the commercial transition aggressively, and let the engineers who built Orion apply their skills to systems that can actually fly at cadence. The Artemis II crew proved humans can return to lunar space. The program that got them there cannot be the program that builds what comes next. $4 billion per launch, one use, no recovery. Rockets do not care about sunk costs, and neither should we.
