Starship at the Crossroads: Can Test 10 Prove Musk Right?
By Miles O’Brien
On Sunday, SpaceX will try again. Another stainless steel behemoth, bristling with 33 engines, will light up the Texas coast and carry the fire skyward. If all goes well, Starship will climb toward orbit, separate cleanly from its booster, and survive the fiery plunge back through Earth’s atmosphere. If it doesn’t, we may see another spectacular explosion - or as SpaceX prefers, a rapid unscheduled disassembly.
There really aren't any euphemisms to sugarcoat the stakes for this flight test. Starship is supposed to be the workhorse not only for Elon Musk’s dream of Mars, but also for NASA’s Artemis program — America’s long-awaited return to the moon. And I can’t help but wonder, as I prepare to watch the test: have Musk and his team designed the future of spaceflight, or a stainless steel dead end?
Despite receiving more than $20 billion in taxpayer funding over the years, SpaceX remains frustratingly opaque about what happens inside its fences. Space enthusiasts and journalists must resort to aiming telephoto lenses at Starbase, parsing hardware movements and launch pad activity like Kremlinologists once did in the Cold War — divining shifts in power by noting which lights were on in the Kremlin at night.
So I got busy calling up some smart people who have flown to space, designed rockets, managed missions, run the space agency, and worked inside the fences of Musky World. Synthesized, their comments offer up an unpolished Starship reality check. So strap in for a pre-flight briefing.
A Culture of Failure — and Its Limits
SpaceX has always thrived on failure. Long before Starship, the company cut its teeth on Falcon 1 — a small, kerosene-fueled rocket launched from a Pacific atoll. The first three launch attempts were failures. Then, in September of 2008, two and a half years after the first attempt, Falcon 1 reached orbit. Its successor, Falcon 9, performed well from the outset, achieving low Earth orbit on its first flight in June of 2010. SpaceX encountered some failures in the quest to fly back and recover the first stage booster. I was there at the Cape in December of 2015 when the first booster landed on terra firma right in front of us. It was a thrilling feat that none of the people I was with predicted would ever happen. Today, they land so routinely that it barely makes the news. The company’s “fail fast, fix fast” mantra rewrote the playbook for rocketry - or more accurately cribbed from the chapter written in the early days of NASA when the agency was less risk-averse, had a lot of young talent, a clear mission, and a blank checkbook.
But Starship is testing that formula to its limits. In recent flights, the same problems have recurred: structural failures, propulsion breakdowns, and shredded heat shields. Usually, each SpaceX failure pushed the ball further down the field. Not this time. For three tests in a row, they’ve stumbled in place.
When rockets were smaller and simpler, each fiery crash revealed a new flaw, and every subsequent launch pushed the boundary a little further. But with Starship — the largest, most complex rocket ever built — the failures are so destructive and multifaceted that clear lessons are harder to extract. At this scale, failure is no longer just a stepping stone; it risks becoming a rut. The very philosophy that made Falcon 9 a triumph may be showing diminishing returns when applied to a rocket of unprecedented size and ambition.
Booster vs. Ship — The Anatomy of Starship
It helps to clarify what “Starship” actually means, because SpaceX uses the name to describe both the whole rocket and just the upper stage.
The full launch system is two parts stacked together:
Super Heavy booster: the massive first stage, 230 feet tall, powered by 33 methane-fueled Raptor engines. Its job is brute force: to muscle the upper stage out of Earth’s dense atmosphere. After separation, it flips around and attempts to return to the launch pad, where giant steel “chopstick” arms are supposed to catch it. It never goes to orbit. Think of it as raw power, meant to be flown again and again.
Starship spacecraft (the “ship”): the upper stage, 165 feet tall, with six engines. This is the payload carrier and deep-space craft. It’s designed to go all the way to orbit, deliver satellites, cargo, or crews, survive reentry using thousands of ceramic heat shield tiles, and then land vertically on Earth. In the future, it could land on the Moon or Mars.
Together, the booster and ship make the tallest, most powerful rocket ever built — 397 feet from bottom to nosecone.
Engines and Vibrations
Starship’s 33 Raptors are there to provide thrust, yes — but also more vibration, resonance, and failure points. The Soviet N1 moon rocket had 30. It never made it to orbit.
Each Raptor is a technical marvel, but the sheer number creates brutal challenges for control and reliability. Every engine’s vibration stresses its neighbors, forcing heavier structures to keep the rocket intact. That erodes payload margins. As you watch the plume of 33 engines ignite, remember: every one of those flames is a potential failure point.
A Workforce Under Strain
Starbase, SpaceX’s test site at Boca Chica, is remote, harsh, and demanding. It’s not an enticing place to live, especially for seasoned engineers with families. Many veterans of Falcon 9 and Dragon stayed behind in California. The result is a younger, less experienced workforce under relentless pressure to deliver.
I’ve heard morale is low. Burnout is real. The stakes are high. Sunday’s test will reveal not just engineering progress, but whether SpaceX’s culture of breakneck iteration is sustainable at this scale.
A successful flight could get the program back on track, but even then, there are huge obstacles that lie ahead.
The Refueling Burden
Perhaps Starship’s biggest Achilles’ heel is fuel. To get to the moon, it must refuel in Earth orbit. Not once, but many times. Depending on who you ask, that could be four tanker flights — or nineteen. Even at the low end, the logistics are staggering. Each refueling requires a flawless docking, transfer of super-cold methane and oxygen, and careful management of boil-off.
No one has ever done this. Not even close. On the ground, cryogenics are hard enough. In orbit, in sunlight and shadow, it’s like juggling dry ice in a sauna.
And that’s just for the moon. For Mars, the challenge multiplies. This is why Musk chose to fuel Starship with methane. In theory, it can be made on Mars using a chemical reaction that combines carbon dioxide with hydrogen to produce methane and water - the Sabatier process. The dream is to send robotic Starships ahead to brew propellant before humans arrive. But as one seasoned hand told me, “I’d want to see a full gas station waiting before I launched.”
Mars Dreams, Earthly Limits
Musk insists Starship is a Mars machine. But studies show a minimal Mars mission would require a vast orbital “transportation node” to fuel up Mars-bound Starships. It would require one to two million metric tons staged in low Earth orbit (roughly two space stations’ worth), capable of storing cryogenic fuel for months or years. Mars still feels like science fiction.
Rockets Are Mostly Fuel — and Why Steel Matters
When it comes to rockets, almost everything is fuel. At liftoff, a typical chemical rocket is about 85–90% propellant, another 8–10% structure, and only a few percent payload. The Saturn V, for example, weighed nearly 3,000 metric tons on the pad, but delivered just 140 tons to orbit — about 5% of its initial mass.
SpaceX’s Starship, stacked atop the Super Heavy booster, will weigh around 5,000 metric tons when fully fueled. Even if it achieves Musk’s original promise of 150 tons to orbit, that’s only 3% payload fraction. If the real performance ends up closer to 50 or 60 tons, it drops to a razor-thin 1%. In rocketry, those margins matter — a small increase in structural weight can erase huge chunks of payload.
That’s why Musk’s decision to build Starship and its booster out of stainless steel was so striking. Steel is heavier than aerospace-grade aluminum or carbon composites. Still, it comes with compensating advantages: it’s cheap, it’s tough, and it keeps its strength in both extremes — the cryogenic chill of liquid methane and oxygen, and the searing heat of atmospheric reentry. For a rocket meant to fly often and be reused quickly, ruggedness and cost can outweigh the penalty in mass.
The gamble is simple: will the benefits of stainless steel — durability, affordability, and manufacturability — be enough to offset the payload performance it gives up? Starship’s shifting numbers suggest that the tradeoff may already be biting. Like the shuttle before it, the weight of making the system reusable could leave less room for the payloads that justify the rocket in the first place.
The Economic Prize
If Starship works, the payoff is enormous. Launch costs could plummet from Falcon 9’s $2,000 per kilogram to mere hundreds. That would change everything: giant space telescopes, orbital power plants, even lunar bases. Musk loves to talk about Mars, but the real prize may be economic — a whole new space economy unlocked by cheap, heavy-lift access.
Yet the demand isn’t there, at least not yet. Even if costs drop, who needs dozens of super-heavy launches a year? Perhaps the most likely customer is SpaceX itself to maintain its massive network of Starlink communication satellites.
The Critical Path
The stakes go far beyond Starship itself. Because NASA has chosen it as the primary lunar lander, the entire Artemis schedule is now chained to SpaceX’s progress. If Starship cannot demonstrate reliable flight, orbital refueling, and lunar descent in time, the timeline for returning astronauts to the Moon will almost certainly slip. Artemis III, the mission that would bring humans to the surface, is nominally slated to launch no earlier than mid-2027. Two years from now? That date also falls into the realm of science fiction.
Meanwhile, China's Space Program is methodically marching toward its own Moon landing. Right now, it appears Chinese taikonauts will be there to greet US astronauts if or when they get there. In that sense, Starship’s testing setbacks aren’t just puncturing Elon's Red Planet vision; they are making the US and also-ran in Moon Race 2.0.
A Buck Rogers Lander
NASA’s choice of Starship as the first Artemis lunar lander is as bold as it is questionable. The vehicle was designed as a multipurpose spacecraft and Mars transport, not a specialized lander. Its sheer size creates practical problems on the Moon: a narrow landing footprint and a towering center of gravity that raise the risk of tipping on uneven terrain, and astronauts must descend more than 100 feet by elevator to reach the surface. The rocket’s engines will also blast out enormous craters of lunar dust and rock during descent, potentially damaging the vehicle or nearby equipment. By contrast, smaller, purpose-built landers like those proposed by competitors would have offered more stability and easier crew access. In betting on Starship, NASA has tied its lunar program to an unproven megastructure that may be ill-suited to the very task it’s been hired to perform.
Safety Compromised
There’s one Starship feature — or omission — that makes me shudder: it has, like the space shuttle, no launch escape system. Unlike Apollo capsules or even SpaceX’s own Crew Dragon, Starship offers astronauts no way out if something goes wrong. In theory, the ship could sometimes separate and limp to safety. But in other phases of flight, there is no escape.
Think about that as you watch the live stream. If people were on board during any of the last few test flights, they would have died. Full stop.
NASA has neatly sidestepped this problem for Artemis. Astronauts won’t ride Starship off the launch pad. Instead, an SLS rocket will loft an Orion capsule that will ferry them to lunar orbit, where they’ll rendezvous with an uncrewed Starship lander. That bureaucratic sleight of hand avoids certifying Starship for ascent. But eventually, Musk intends to put as many as 100 people on top of that booster. I wouldn’t volunteer for the first seat.
My Take on the Flight 10
So what should we expect on this next test? If the booster ignites cleanly, if separation works, if the ship’s heat shield holds, SpaceX will claim victory. And it will be progress. Every test, even the failures, generates terabytes of data that engineers sift through with AI-driven analysis. That’s one of SpaceX’s real strengths.
At some point, failure must yield progress. If it doesn’t, the whole philosophy starts to look more like recklessness than innovation. Sunday’s test may be the moment we find out if SpaceX is still learning, or simply repeating mistakes.
And if the same problems recur — tiles shedding, engines failing, structural anomalies — it will confirm my unease that Starship has hit a wall. Musk has a way of proving doubters wrong. He did it with Falcon 1 and Falcon 9’s booster landings. But this time the challenges are an order of magnitude greater.
I’ll be watching with a mix of awe and skepticism. Awe at the spectacle of 33 engines lighting at once. Skepticism about whether this grand stainless steel gamble can deliver more than fireworks.
Starship could ignite a new era of exploration — or expose the limits of Elon Musk’s audacity. This launch may not give us the final answer. But it will be another data point in the most ambitious, most perilous, and most fascinating rocket program of our time.
###











Thanks!
What genius at NASA thought this approach was a good idea? Going with the head DOGE guy is more than reckless and stupid. It's homicidal. Great report, Mr. O'Brien. Love your work.