Tesorb Signal: Tesla, SpaceX, and the Musk Companies

There’s a plan on file with Clark County, Nevada that describes a transit system with one hundred four stations. Underground Teslas gliding beneath the Strip, the stadium, the airport, and downtown.

There’s a plan on file with Clark County, Nevada that describes a transit system with one hundred four stations. Underground Teslas gliding beneath the Strip, the stadium, the airport, and downtown. Here’s how many of those stations you can actually ride to today. About a dozen. Five years in, that gap is the whole story.

This is Tesorb Signal and I’m Lena Ruiz. Today, our first proper look at The Boring Company. Three things. What the Vegas Loop actually is in twenty twenty-six. The promise on file versus the pace on the ground. And the two milestones that decide whether any of this scales.

The origin story is pure Musk. The company began a decade ago as a half joke about being stuck in traffic, a vow to build a tunnel machine and start digging. Las Vegas became the proving ground, and a system running under the convention center by twenty twenty-one became the receipt.

First, credit where it’s due, because the core product works. Since twenty twenty-one, the Loop has carried more than four million passengers through its tunnels.

The trick is almost funny in its simplicity. Regular Teslas, driven by human drivers, rolling through concrete tubes about forty feet down. A ride across the convention center campus takes a couple of minutes instead of a twenty-five minute walk.

On its busiest convention days, the system has moved tens of thousands of riders in a single day. And the network keeps inching outward. The original convention center loop grew into stations at Strip resorts, Resorts World, Westgate, Encore, and most recently Sahara. There’s even airport service of a sort, with a small number of daily rides now reaching the terminals at Harry Reid International.

The construction speed claims also hold up better than a skeptic might expect. The tunnel connecting Encore to the convention center went from site access to completed tube in under ten weeks. Digging tunnels cheaply and quickly was the founding pitch of this company, and on that narrow question, it has receipts.

Now put all of that against the plan on file. Clark County has approved a network wrapping the airport, the stadium, and downtown into one underground grid. In its final form, the company says the system would move up to ninety thousand passengers per hour.

For scale, the full station map is bigger than the entire Washington Metro. That is not a shuttle. That’s a subway sized ambition, built out of Teslas.

And every version of that ambition runs through the airport. Tens of millions of travelers move through Harry Reid every year, and Las Vegas officials have called a direct airport link the holy grail of local transit.

But today’s airport rides include a long stretch of ordinary surface driving. The fix is a pair of twin tunnels called the Airport Connector, and that’s where the timetable starts to slip.

The company’s own target for those tunnels was hopeful and specific. Open in the first quarter of this year. It’s August, and the connector is still under construction.

Meanwhile, talks are only now underway about extending the Loop into the airport’s actual footprint, rather than stopping at its curb.

And the stakes of that opening are not small. By the company’s own projections, a working airport link could multiply the system’s entire annual ridership several times over. The connector isn’t one more station. It’s the difference between an attraction and infrastructure.

There’s a similar rhythm in daily operations. For most of its life, the system has mostly run when the convention center runs, and the free rides on campus give way to paid tickets beyond it, a few dollars to fifteen depending on distance.

None of that is damning on its own. What exists today is a genuinely clever amenity, and a well liked one. But what was promised is a utility, a system a city leans on every hour of every day. Those are different products, and only one of them justifies the map on file.

Zoom out and the pattern holds elsewhere. A Loop is under construction in Nashville, where the progress is real and so is the ordinary friction that meets every transit project.

A first international project is signed in Dubai. Each announcement arrives with a bigger map than the last. The building really does happen. It just happens at a fraction of the speed of the announcing. After five years, roughly one station in nine from the Vegas plan is open.

There’s also a quiet Tesla story tucked inside this one. The Loop is the most natural home for the Cybercab, a closed track with no weather, no cyclists, and no cross traffic, a thread we’ve been tracking since our tunneling deep dives.

When driverless cabs eventually replace the human drivers down there, the Loop’s economics change completely. The tunnels may end up mattering less than what drives through them.

So the honest scorecard reads like this. The tunnel works. The cars work. Four million rides say the product is real. What’s unproven is the multiplication. Getting from a dozen stations to the full map isn’t an engineering problem. It’s a permitting, funding, and construction cadence problem, and five years of evidence says that cadence runs slow.

So the tunnel isn’t the test. The timetable is. Watch two dates. The day the Airport Connector actually opens, and the day a Cybercab takes its first paying tunnel ride with nobody up front. Either one would turn the biggest map in Las Vegas from a filing into a schedule. When they happen, you’ll hear it here.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. Corrections are posted to the episode page.

After Flight 13’s near-perfect ocean splashdown, SpaceX is going for the real thing: catching Starship’s upper stage with the launch tower arms. Flight 14 will also be the first to reach orbit and carry real Starlink cargo. If it works, the test program is over.

For days after its last test flight, a stainless steel spaceship longer than an Olympic swimming pool floated in the Indian Ocean like a message in a bottle. It had landed soft, upright, and almost exactly on target. SpaceX looked at the numbers and made a call. The next one doesn’t get to swim. The next one comes home.

This is Tesorb Signal and I’m Lena Ruiz. Today, the flight that could end Starship’s test era. Three things. What Flight fourteen will attempt. Why catching a ship changes the economics of space. And what happens if it goes wrong.

Some quick table setting. Last month, Starship’s thirteenth test flight sent its upper stage most of the way around the planet and brought it down to a gentle splashdown, the flight we covered when it happened. After six days of data review, Musk confirmed the landing was so precise that the tower arms would have caught it. That one sentence unlocked everything that comes next.

The plan for Flight fourteen came out on SpaceX’s first earnings call as a public company, and it’s a stack of firsts. The first Starship flight to actually reach orbit. The first attempt in history to catch an orbital spacecraft out of the sky with a launch tower. And the first flight carrying real cargo, a batch of next generation Starlink satellites. The target is late August.

Announcing it that way carries its own weight. A date and a promise, delivered to brand new public shareholders on the company’s first call. If the flight slips or the catch misses, that miss is public now too.

Hold on to the orbital part, because it’s bigger than it sounds. All thirteen previous Starship flights stayed suborbital on purpose. A ship that has a problem on a suborbital arc falls into the ocean no matter what.

A ship that reaches orbit and can’t come down properly is fifty meters of stainless steel with no guaranteed place to land. Going orbital means trusting the vehicle in a whole new way.

Then comes the showstopper. Since late twenty twenty-four, SpaceX has repeatedly snagged the Super Heavy booster out of the sky with the tower’s chopstick arms, then refurbished some of those boosters and flown them again.

The ship is a different animal. It returns from orbital speed, scorched from reentry, and has to thread itself between two metal arms. No one has ever recovered an orbital class vehicle this way.

One detail worth noting. On this flight, only the ship gets the catch attempt. The booster that lifts it will splash down in the Gulf, the way it has on recent flights. SpaceX is isolating the new trick and changing one variable at a time. That’s the same discipline that turned booster catches from a spectacle into a routine.

The catch itself will look like physics showing off. The ship falls belly first, flips upright, relights its engines, and slows to a hover beside the tower while the arms close under hardware points on its sides. Every one of those steps has now worked over open water. Doing them beside steel and concrete is the exam.

So why go through all this instead of fishing ships out of the sea? Because the whole point of Starship is full and rapid reusability, and the money only works if both halves come home ready to fly again.

Ocean recovery means salt water, recovery ships, cranes, and weeks of delay. Tower recovery means the vehicle lands next to the pad that launches it. The stated goal has always been airline style turnaround.

For proof of what reuse does at scale, look at SpaceX’s workhorse. Falcon nine just flew its ninetieth mission of this year, with the ninety-first launching hours later. That’s a rocket flying more than twice a week, and it only reuses its bottom half. A caught and reflown Starship is that same math applied to the biggest flying machine ever built.

The payload explains the urgency. Those next generation Starlink satellites are too big for Falcon nine. Scaling the network that pays SpaceX’s bills depends on Starship flying often, which depends on catching and reflying ships. NASA is watching just as closely, because this same vehicle is supposed to land astronauts on the Moon.

To be fair about the odds, Flight thirteen already proved the hard parts once. The heat shield held, the flaps steered through reentry, and the ship hit its mark in the open ocean. SpaceX has also caught boosters on the first try before. This isn’t a company guessing. It’s a company repeating a trick one size up.

But the schedule pressure is real and documented. A federal watchdog reported last month that SpaceX is running more than a year behind its original timeline on the Moon lander program. The catch attempt still needs regulatory sign off. And a miss over land, next to the launch pad, costs more than a miss over water ever did. The ocean forgives. The tower doesn’t.

Step back and the stakes get simple. A rocket you fish out of the ocean is an experiment. A rocket you catch, refurbish, and fly again is a product. Thirteen flights built the experiment. Flight fourteen is the audition for the product. And after ninety Falcon missions in a single year, nobody can call the business model behind it theoretical.

So this is the flight where the test program either graduates or repeats the grade. Watch the final minute, not the first. If those arms close around a ship that left the planet and came back, the era of throwing rockets away is effectively over. Either way, you’ll hear the verdict here.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. Corrections are posted to the episode page.

Grimes County, Texas just signed up to host the largest building on Earth. SpaceX’s Terafab chip plant will span 100 million square feet of manufacturing space. But almost every detail has changed since the project was first announced four months ago, and the deal was done before the county got to ask questions.

Grimes County, Texas has more cattle than people. Last week, it signed up to host the largest building on Earth. Not the largest factory in America. The largest building our species has ever put up. And the company behind it confirmed the deal was already done before most of the county got to ask a single question.

This is Tesorb Signal and I’m Lena Ruiz. Terafab has an address now. Today, three things. What actually got announced last week. What quietly changed since the version we covered in the spring. And what the deal looks like from the county that just signed it.

Start with the announcement. Last Thursday, Texas Governor Greg Abbott confirmed that SpaceX will build its Terafab chip plant in Grimes County, northwest of Houston. One site handling the whole chip pipeline. Logic. Memory. Advanced packaging. All of it under a single roof.

The rollout came in stages. SpaceX representatives confirmed the project at a county meeting on Wednesday. The governor made it official on Thursday. And SpaceX posted a first rendering the same day. A single structure laid across ranch land, stretching roughly two and a half miles from end to end.

And the size is the headline. The plans call for more than one hundred million square feet of manufacturing space. For scale, Giga Texas in Austin covers about ten million square feet, roughly a hundred football fields of floor, and it’s already one of the largest buildings on Earth. Terafab would be ten of those. In one building.

Musk calls it the largest and most valuable building on Earth by far. The chips it makes are meant for three customers you already know. Optimus robots. Cybercabs. And the data centers SpaceX wants to operate in orbit. Chips for use on Earth and in space, in the company’s own words.

The first phase carries a price tag of sixteen point eight billion dollars and a promise of three thousand permanent jobs, ranging from food service to advanced chip engineering. The state sweetened the deal with a grant and a spot in its tax incentive program for giant capital projects.

Now rewind four months, because long-time listeners have heard this name before. Terafab was the very first story this show ever covered. Back then, it was a twenty-five billion dollar fab planned for the Giga Texas campus in Austin, with Intel signed on as the manufacturing partner and a target of one terawatt of AI compute per year.

Almost every detail from that spring version has moved. The site left Austin entirely. The budget changed shape, from one number for the whole project to a first phase with no public figure for the final total. And Intel, the partner that once anchored the whole story, barely rated a mention last week, though local reporting still lists it as part of the proposal.

To be fair, there’s honest math behind the move. Musk himself admitted the Terafab couldn’t possibly fit on the Giga Texas campus. Austin keeps a smaller research fab while the full-scale plant needed open land, and Grimes County has plenty. Big projects relocate during planning all the time. On its own, that part is normal.

The strategic case is real too. America learned during the pandemic what happens when the chip supply lives somewhere else. Tesla and SpaceX both argue their future demand for AI chips will outrun what the world’s foundries can deliver. If you believe that forecast, building your own supply isn’t vanity. It’s survival.

At the county meeting, a SpaceX lead went further, calling the project strategically vital for the United States.

But the view from Grimes County is more complicated. On Wednesday, hundreds of residents packed the county commissioners court for the first public question and answer session with SpaceX. And one of the first things they learned was that the agreements were already signed and active. The deal was done before the meeting started.

The money helps explain the tension. Between the county and local school districts, the tax breaks add up to an estimated seven hundred ten million dollars over thirty-five years. Residents pushed back on the size of those breaks, and on how little they were told while the terms came together.

There’s another side to that story too. The local school boards that approved the agreements say the project will strengthen their districts, and the county judge called it a generational change. For a lot of neighbors in a rural county, this factory is the biggest economic news in living memory. Their bet is that the jobs and the tax money outweigh the growing pains.

SpaceX’s answers came in two parts. On taxes, it projects that even with the breaks, its yearly bill at peak will rank among the largest any company pays in the state. On water, it says the plant will draw from the Gibbons Creek Reservoir instead of local groundwater. Real answers. They just arrived after the signatures did.

Step back and look at the arc. In four months, Terafab changed its address. Then its budget structure. Then its scale. What never changed was the superlative. Biggest. Each revision made the claim grander while the specifics got harder to pin down. One hundred million square feet is a stunning number. Today, it’s also just a rendering and a stack of signed agreements.

So keep two ledgers on this one. One for what gets announced, and one for what gets built. The announcement ledger grew again last week. The construction ledger is still mostly empty ranch land in Grimes County. When steel starts rising northwest of Houston, you’ll hear about it on this show.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. Corrections are posted to the episode page.

Tesla is about to re-reveal a car it first showed in 2017. The Roadster, with its SpaceX cold gas thrusters, $50,000 deposits, and nine years of broken dates, is finally getting a stage in Texas. The question is what all that waiting actually bought.

Sometime this month, if the reports hold, Tesla will unveil a car. Not a new car, exactly. A car it first unveiled back in twenty seventeen. A car customers paid fifty thousand dollars to reserve. A car that was supposed to ship in twenty twenty. The Roadster is finally getting its second reveal. The open question is what nine years of waiting actually bought.

This is Tesorb Signal and this is Lena Ruiz. Today, the car Tesla promised and why it’s genuinely hard to build. Then, the promise ledger, date by date. And finally, what those deposits say about the strange economics of trust.

Start with the car itself, because the ambition is real. The Roadster Tesla showed in twenty seventeen was pitched as the fastest production car ever built. Tesla’s website still lists a top speed above two hundred fifty miles an hour. And the headline feature has only gotten wilder since.

That feature is a cold gas thruster system developed with SpaceX, known inside the company as A71. Compressed air, fired through nozzles around the body, meant to sharpen acceleration and braking and cornering. An early version was demonstrated to Musk himself this spring. He has promised the most exciting product demo ever. In his words, whether it works or not.

Tesla has hinted the thrusters could do more than grip the road. Musk has teased hover capability for years, a party trick borrowed from his rocket company. Whether that survives contact with regulators and physics is anyone’s guess. But it explains why he keeps calling this event a demo, not just a launch.

And in fairness, nobody has ever shipped a production car with thrusters. This is genuinely new engineering, and genuinely new engineering rarely runs on schedule. If the system works, the Roadster stops being a car company’s halo project and becomes something closer to a physics demonstration you can register for the road.

A second point of fairness. Tesla spent the past decade scaling the Model Y into the world’s best selling car and pouring billions into autonomy. Choosing volume and software over a hand-built halo car is a defensible business call. And the Roadster deposits are refundable. Some frustrated reservation holders have asked for their money back, and they got it.

Now open the ledger. Twenty seventeen. Tesla rolls the second generation Roadster onto a stage next to the Semi truck and promises production in twenty twenty. Reservation holders put down five thousand dollars up front, then another forty-five thousand within ten days to lock a build slot.

Twenty twenty comes and goes with no Roadster. Nearly every year after, a new date gets floated and missed. In early twenty twenty-four, Musk says the car will be shown by the end of that year and ship the following year. It isn’t, and it doesn’t. Then last November, at Tesla’s shareholder meeting, he finally gets specific.

The reveal, Musk announces, will happen on April first, twenty twenty-six. He jokes that he’s keeping some deniability, since that date happens to be April Fools’ Day. April first arrives. No reveal. The target slides to late April. On the first quarter earnings call, Musk says the unveil is about a month away. Then it’s May. Then early June.

Now, according to a report from The Information, a tech publication, the event is set for this August, in Texas. Tesla hasn’t confirmed the date, and hasn’t denied it. If it holds, this will be Tesla’s first vehicle debut since the Cybercab, almost two years ago. And production is now targeted for around mid twenty twenty-seven, at Giga Texas.

The location matters too. Texas is where the car will be built, and Texas is where Tesla’s story increasingly lives. Its headquarters. Its biggest factory. Its robotaxi experiment down in Austin. The Roadster reveal doubles as a statement about where the company’s center of gravity has moved.

Do the arithmetic on that timeline. A car revealed in twenty seventeen, re-revealed in twenty twenty-six, entering production in twenty twenty-seven, reaching customers as much as a decade after the first stage lights came on. In that same stretch of time, Tesla designed, launched, and scaled the entire Model Y program from nothing to the best selling car on Earth.

We walked through the Roadster’s broken promises on this show earlier this year, back when the count stood at nine years. The count is still nine years. What changed is that the dates now move in months instead of years, which is either progress or practice, depending on your patience.

So think about what those deposits mean. A customer who reserved in twenty seventeen has had five figures parked with Tesla for almost nine years. That money earned Tesla interest instead of the customer. In exchange, the customer has received a rotating series of dates, each one quietly replaced by the next. The car may still turn out to be spectacular. The wait already tells its own story.

One comparison makes the point. When Tesla launched the Cybertruck, reservations cost one hundred dollars, and the crowd treated missed dates as part of the show. The Roadster asked for five hundred times more, from customers who expected to be first in line. Patience scales differently when the deposit has a comma in it.

So bring the receipt to the reveal. When the sheet comes off in Texas, the demo will be judged in seconds. Thrusters firing. Numbers on a screen. A crowd deciding in real time whether the wait was worth it. But the fifty thousand dollar receipt sitting in a reservation holder’s drawer took nine years to age. Both of those things are the product. Watch the demo. Remember the receipt.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. Corrections are posted to the episode page.

On Tesla’s latest earnings call, Elon Musk explained why a robotaxi needs a satellite dish. But that Starlink dish on the Cybercab is one wire in a much larger bundle connecting Tesla’s balance sheet to SpaceX, including a $1 billion mark-to-market gain that made up most of Tesla’s Q2 profit.

On Tesla’s latest earnings call, Elon Musk explained why a robotaxi needs a satellite dish. Driverless cars, he argued, can’t be getting stranded in dead zones where cell service disappears. Fair enough. But that dish on the Cybercab is one wire in a much larger bundle. Pull on it, and you can trace connections running all the way from Tesla’s balance sheet to a rocket company.

This is Tesorb Signal and this is Lena Ruiz. Today, why Starlink in a robotaxi genuinely makes sense. Then, the money already flowing between Musk’s two biggest companies. And finally, the number buried in Tesla’s profit line that deserved far more attention than it got.

Start with the engineering case, because it’s real. Musk told analysts that a robotaxi fleet can’t afford connectivity dead zones. He compared them to Bermuda triangles, places where a driverless car could effectively go missing.

A car with no driver and no signal is a car nobody can find, reroute, or rescue. Starlink works anywhere it can see the sky. For a fleet with no humans aboard, that’s not a luxury feature. That’s infrastructure.

The passenger case follows naturally. People riding in a driverless car want to work, or watch something, or take a call. Musk says Starlink can stream live sports in 4K at a cost per gigabyte that cellular networks can’t match. So the plan is Starlink in the Cybercab first, then across the rest of Tesla’s lineup in markets where the service operates.

And Starlink can back up the promise, because it’s no longer a side project. Twelve million subscribers, double what it had a year ago. Four point two nine billion dollars in revenue in a single quarter. The dish has become one of the most valuable internet businesses on the planet. Putting it in a car is not a stretch. It’s a product decision.

One more point of fairness before we go deeper. Everything you’re about to hear comes from public disclosures. The equity stake sits in regulatory filings. The customer relationships were itemized in SpaceX’s IPO paperwork. The quotes come from recorded earnings calls. Whatever you conclude about this entanglement, it is happening on the record, in the open.

Now trace the wire. Tesla put two billion dollars into xAI. When SpaceX acquired xAI early this year, that investment converted into a small equity stake in SpaceX itself. It was the first time a regulatory filing formally connected the two companies. It was not the last.

The IPO paperwork showed Tesla has sold roughly eight hundred ninety million dollars worth of vehicles and batteries to SpaceX and xAI since twenty twenty-three. SpaceX buys Cybertrucks from Tesla. It buys battery packs. And a fresh filing added yet another purchase to the pile.

Then add the parts that don’t show up as line items. Grok, the xAI chatbot, runs inside Tesla’s cars. Starlink is headed into the Cybercab. And Terafab, the giant chip project in Austin, is being built as a shared foundation for both companies’ AI ambitions. On the earnings call, an analyst asked Musk about all this overlap. His answer went further than the question.

Musk said there’s more and more overlap with SpaceX, especially on Terafab. Then he added that combining companies isn’t something you can talk about on an earnings call, because it has to be done through the appropriate process. Sit with that phrasing for a second. He didn’t confirm a merger. He didn’t deny one. He described the procedure for one.

Shareholders were already asking. Twenty-two separate questions about a possible Tesla and SpaceX combination made the formal agenda for that call. We’ve covered before how SpaceX keeps showing up in Tesla’s books as one of its best customers. The relationship has now moved a step past customer. Tesla is a shareholder.

If this feels hard to picture, think of two houses on the same street slowly knocking doorways through their shared wall. Each doorway has a reasonable purpose. A chatbot here. A satellite dish there. A chip factory in the middle. No single doorway merges the houses. But walk through enough of them, and the floor plan is one building.

Which brings us to the profit line. Tesla reported about one point one billion dollars in net income for the second quarter. Sitting inside that number is a mark-to-market gain of one billion dollars on Tesla’s SpaceX stake. That’s an accounting entry. SpaceX’s stock went up, so Tesla’s books recorded a profit. No car was sold to earn it.

Strip out the paper gain and look at operations. Tesla’s income from actually running its business was three hundred ninety-eight million dollars, an operating margin of about one point four percent. Razor thin. Read plainly, most of the quarter’s reported profit came from owning a piece of another Elon Musk company. Not from selling cars. Not from selling energy. From holding stock.

To be clear, mark-to-market accounting is standard practice, and Tesla disclosed exactly where the gain came from. The number isn’t hidden. It’s just doing a lot of work. An operating margin that thin is the kind of result that once triggered crisis headlines. This quarter, it sat quietly underneath a healthy looking profit.

The satellite dish is the smallest part of this story. Starlink in a robotaxi is sensible engineering, and on its own it would barely be news. But every quarter, another wire gets soldered between these two companies. Vehicles. Batteries. Chips. Chatbots.

And now a one billion dollar paper gain sitting where the profit used to be. Whether or not a merger ever gets announced, the filings tell you the finances are already merging on their own.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. Corrections are posted to the episode page.

A Diamond Black Model Y rolled off the Fremont line last week as Tesla’s 10 millionth car. No EV-only automaker has hit that mark. But the milestone masks a tougher story: two years of falling deliveries, market share below 50%, and the second half of Musk’s pay package looking steeper than the first.

Last week, a Model Y painted Diamond Black rolled off the line in Fremont, California, and factory workers popped confetti over the hood. It was the ten millionth Tesla ever built. No company that makes only electric cars has ever done that. So why did the celebration feel less like a finish line, and more like a mile marker on a much harder road?

This is Tesorb Signal and this is Lena Ruiz. Today, the milestone and how fast it arrived. Then, the sales numbers the confetti doesn’t show. And finally, the one person with about a trillion reasons to care whether the second half of this journey ever gets finished.

Start with the number itself, because it has earned a moment. Tesla needed about twelve years to build its first million cars. That millionth car rolled out of Fremont in the spring of twenty twenty. The next nine million took just six years. Same company. Ten times the output. A fraction of the time.

For scale, consider what that pace means. Across those six years, Tesla’s factories averaged roughly a million and a half cars a year. That’s a new Tesla rolling off a line somewhere in the world about every twenty seconds, around the clock, for six straight years.

And Tesla did it with only four assembly plants. Fremont in California. Giga Shanghai. Giga Texas. And Giga Berlin. Fremont itself carries some history. General Motors and Toyota once ran that plant as a joint venture, and Tesla bought most of it in twenty ten for about forty-two million dollars. The auto industry’s castoff became the birthplace of the milestone car.

Credit where it’s due. Tesla is the first company built only on electric cars to reach eight digits of production. BYD reached the same count earlier, but that total includes plug-in hybrids with gas engines. For pure battery EVs, Tesla got there first. That is real execution, not a marketing line.

And a second thing worth crediting. Tesla built this scale while the loudest voices in the industry spent years calling electric cars a niche, a fad, or a money pit. A decade ago, the debate was whether EVs could work at all. That debate is over, and this factory in Fremont is a big part of the reason why.

Now set the confetti down and pull up the sales chart, because it tells a different story. Tesla’s best year was twenty twenty-three, when it delivered a record one point eight one million cars. Since then, deliveries have fallen two years in a row. Last year alone, sales dropped about nine percent.

The sharper number is market share at home. Back in twenty nineteen, nearly eight out of every ten EVs sold in America were Teslas. For all of twenty twenty-five, Tesla’s share of the American EV market was about forty-four percent. That was the first full year it fell below half. The pie kept growing. Tesla’s slice kept shrinking.

And the competition taking those points isn’t one rival. It’s everyone at once. Legacy automakers finally shipping credible EVs. Korean brands undercutting on price. And Chinese manufacturers scaling faster than almost anyone predicted. Tesla didn’t get worse at making cars. The rest of the field got better.

We’ve talked before on this show about BYD taking the global EV sales crown from Tesla. This is that same story, told from a different angle. Think of it like the dashboard of a car. A production milestone is the odometer. It tells you how far you’ve traveled. Market share is the speedometer. It tells you how fast you’re moving right now.

Tesla’s odometer just hit a number no other electric carmaker has touched. The speedometer is the problem. Wall Street noticed the same tension. Tesla’s stock has spent the year drifting lower, even after a delivery beat last quarter. And there’s one person watching both gauges more closely than anyone else.

Elon Musk’s pay package, the one Tesla shareholders approved last year, is worth roughly a trillion dollars in stock if the company hits a series of ambitious targets. One of those targets is twenty million vehicles delivered. The milestone at Fremont means Tesla is, on production at least, halfway to that number.

That package was controversial precisely because its targets sounded out of reach. The Fremont milestone is the first time one of them has a visible halfway marker under it. Which makes the celebration a progress report on the largest pay deal in corporate history.

But the second half is the steep half. Analysts point out that the road to twenty million looks difficult without new high volume models. The lineup that built the first half is aging. The affordable next generation car keeps sliding down the priority list, behind robotaxis and humanoid robots. The math of the next decade is not the math of the last one.

And that’s worth being precise about. Getting from one million to where Tesla stands today took six years of flat out execution in a market Tesla mostly had to itself. Doubling from today, after two straight years of falling deliveries and with rivals multiplying, is a very different problem. Something has to change. A new model. A new market. Or both.

So hold the confetti up against the sales chart. Ten million cars proves Tesla won the last decade of electric vehicles. Nobody can take that away. But a milestone looks backward, and a market only looks forward. The celebration is for cars already built. The open question is who builds the second half, and how long it takes.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. Corrections are posted to the episode page.

Investors are trading Neuralink shares at prices valuing the company at $42 billion, more than four times its last official $9 billion price tag. But between those two numbers, the clinical reality tells a quieter story: 21 participants, no commercial approval, and revenue years away.

In recent weeks, investors have been trading shares of Neuralink at prices that value the company as high as forty-two billion dollars. Its last official price tag, set when it raised money about a year ago, was roughly nine billion. Between those two numbers, a handful of new patients received implants. Nothing about the science changed more than four-fold. Something else did.

This is Tesorb Signal and I’m Lena Ruiz. Three things today. Where the new number comes from, and why it isn’t official. What Neuralink has actually done in the clinic this year. And what a price like this really measures.

First, the mechanics, because they matter. Neuralink hasn’t raised new money at this price, and the company hasn’t confirmed any of it. These are secondary trades. Employees and early backers selling their shares to investors who want in.

According to reporting from Business Insider, recent deals have ranged from twenty-nine billion up to that forty-two billion mark, with some bidders reportedly offering prices that would imply nearly sixty billion.

Notice the spread. When trades in the same company land anywhere across a two-x range, that isn’t a market reaching a verdict. It’s an auction where nobody agrees on what’s being sold. Think of guessing a neighborhood’s home values from a couple of houses that sold quietly, off market. The prices are real. The signal is thin.

For the record, the last investment Neuralink actually disclosed came in May, when Oman’s sovereign wealth fund confirmed a stake, at a valuation neither side shared. Everything since is other people trading the shares among themselves.

What’s driving it is no mystery. The SpaceX stock market debut in June minted enormous paper gains for people who got in early, and it lit a fire under demand for everything else Elon Musk owns that you can’t buy yet. One venture investor told reporters that clients were asking for Neuralink shares at any price. Any price is a mood, not a valuation.

To be fair, the science is not standing still. Neuralink is running trials in multiple countries. This year its new surgical robot placed the implant’s threads directly through the brain’s protective outer layer for the first time, a step toward the faster, more automated surgery the company needs.

A study aimed at restoring speech has a regulatory fast track. And the vision implant, meant to restore sight to blind patients, is targeted for its first human test.

The patients are real too. We spent a full episode with them earlier this year, people running their digital lives through these implants for hours every day. Whatever the price of the shares, that progress is genuine medicine, moving at the careful pace brain surgery demands.

And the buyers aren’t fools. Betting early on Musk companies has paid, repeatedly. Investors in his social platform and his AI company ended up holding SpaceX shares through cross-company deals. Backers of the rocket company got the debut of the decade. If you believe the pattern holds, paying up for the last private door in the building is a strategy, not a delusion.

But now put the two sets of numbers side by side. On the demand side, forty-two billion dollars. On the clinical side, the company’s own January update counted twenty-one participants. There is no commercially approved product. In the United States, Neuralink is still a clinical trial. And as we covered earlier this year, a simpler Chinese implant crossed the commercial approval line first.

Even the friendliest projections, reported from the company’s own internal figures, don’t see revenue arriving in real size until the end of the decade. Whatever these shares earn their new owners, it won’t come from patients any time soon.

And to be clear about what would change that. A commercial approval in the United States. A surgery pipeline running at hundreds of patients a year instead of dozens. A priced funding round where the company itself sets the number. Any of those would be news about Neuralink. A secondary trade is news about the buyer.

There’s a wrinkle worth sitting with. The SpaceX debut that lit this fire has itself cooled. The stock has slid back below its offering price. The appetite for Musk exposure survived that just fine. Which tells you the appetite isn’t tracking outcomes very closely. It’s tracking access.

A price like this is a measurement. The question is what it measures. A public stock price, for all its madness, gets voted on by millions of people every day. A thin secondary market in a private company measures something narrower. How badly a small number of wealthy investors want in the door.

Both are real prices. Only one of them is evidence about the company.

So measure the appetite in dollars. Measure the progress in patients.

The forty-two billion dollar number is a fact about investors, not about brains. The facts about brains are quieter. Twenty-one people, careful surgeries, trials grinding forward in multiple countries, a speech study and a vision study waiting their turn. That ledger will decide whether Neuralink becomes medicine at scale, and it moves at the speed of hospitals, not auctions.

When the next eye-watering Neuralink number crosses your feed, ask one question first. Is that a patient number, or an appetite number? They’re written in the same font. They mean completely different things.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. Corrections are posted to the episode page.

SpaceX flew the best Starship test flight it has ever flown. Real payload (20 Starlink V3 satellites deployed), first engine relight in space, and a ship that survived reentry and floated intact. But 13 engines were commanded to relight for the booster’s landing burn. By published counts, 5 lit. This was also the first Starship test graded by shareholders. Engineers and investors are scoring the same flight differently.

Days ago, SpaceX flew the best Starship test flight it has ever flown. The company’s own commentator called the ending a dream scenario. And buried inside that dream is one number that explains why the hardest part is still ahead. Thirteen engines were commanded to relight for the booster’s landing burn. By published counts, five of them lit.

This is Tesorb Signal and I’m Lena Ruiz. Three things today. What Flight thirteen actually achieved, and it’s a real list. The gap the flight quietly measured. And why this was the first Starship test ever graded by shareholders.

Start with the achievements, because they’re substantial. For the first time in thirteen flights, Starship carried a real payload. Twenty next generation Starlink satellites, pushed out through the ship’s slot door one by one, like cards from a dealer. Every one of them deployed cleanly. Until now, every Starship has flown empty or carried dummy mass. This one did a job.

And not a token one. Those satellites are the new, heavier generation that the entire next phase of Starlink depends on, and Starship is the only rocket built to carry them.

Then the ship did something arguably more important. It relit one of its Raptor engines in space, a short burn, on command. That sounds small. It isn’t. An engine that restarts in space is the difference between a rocket that can enter orbit and leave it, and a rocket that can’t safely go at all. The last flight skipped this test. This one passed it.

And the ending made the highlight reel. The ship survived reentry, steered itself through a banking maneuver, and set down in the Indian Ocean so gently that it stayed intact, floating, and still transmitting. Cameras got the first clear look ever at a heat shield that had just survived the fire. Elon Musk said the flight returned more heat shield data than the team expected.

When the first version three rocket flew back in the spring, we argued that everyone was scoring it wrong, that the data mattered more than the landing. Flight thirteen is that argument paying out. The booster tumble that ended the spring flight was diagnosed, fixed, and flown again inside two months. That’s the iteration machine working as designed.

Now the gap. Rewind to six and a half minutes after liftoff. The booster, coming home, commanded thirteen engines to relight for its landing burn. Most stayed dark. By published counts, five lit. The booster hit the Gulf hard, carrying far more speed than planned.

To be fair, that booster was always headed for the water. SpaceX practices these landings at sea, on hardware it plans to throw away, precisely so failures cost nothing but data. That’s the responsible way to learn.

But the lesson in the data is uncomfortable. One industry analyst put it plainly. If a relight failure like that happened during a real landing attempt at the launch site, the booster could wreck the pad it needs for the next flight.

Engines that restart reliably, every time, are the entire foundation of a rapidly reusable rocket. Right now, that reliability isn’t there.

And step back further. Thirteen test flights. Three years of launches. A decade since the rocket was first unveiled. Starship has still never completed a single full orbit of the Earth.

The clock that cares most about that belongs to NASA. A version of Starship is supposed to dock with another spacecraft next year, and land astronauts on the Moon in twenty twenty-eight. Before any of that, Starship has to transfer fuel between two ships in orbit, a demonstration still waiting on the calendar.

Which brings us to the new scorekeeper. This was the first Starship test since SpaceX went public in June. In the weeks after the debut, the stock gave back all of its early gains and then some, sinking below its offering price. A test flight is no longer just an engineering milestone. It moves a price that ordinary people now own in their retirement accounts.

We covered how index funds swept SpaceX into everyday portfolios. This flight was the first time those new owners watched a launch with money on the table. And the company has since faced Wall Street’s other ritual, its first quarterly earnings report.

Engineers grade a test flight on what was learned. Markets grade it on distance to revenue.

For Starship, revenue at scale means routine orbital flights, sixty satellites at a time, boosters caught and reflown in days. Judged as a test, Flight thirteen was a triumph. Judged as transportation, it was a reminder that the destination is still far away. Both scores are honest. They’re just measuring different things.

So score the flight twice. Once as an engineer, once as a shareholder.

The engineer’s scorecard says Flight thirteen was the best Starship has ever flown. Real payload. Relight proven in space. A heat shield that came home intact. The shareholder’s scorecard says the machine that makes the business work, a booster that restarts every engine, every time, on demand, isn’t finished.

Watch the relight numbers on the next flight. Not the splashdown footage, not the adjectives. The moment those engine counts hit thirteen for thirteen, this stops being a test program and starts being a transportation system. That’s the number that closes the gap between the dream scenario and the business plan.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

Got a tip or feedback? Send it to [email protected], we’d love to hear from you. This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. See you on the next one.

Tesla just posted $28.2 billion in revenue, up 26% year-over-year, its biggest quarter ever. The market erased $140 billion in a single day. EPS came in at $0.33 vs $0.53 expected, a 39% miss. Operating margin: 1.4%, down from 4.1%. This episode covers the record quarter, the miss that buried it, and why the market is done paying on faith.

Tesla just posted the biggest quarter in company history. Twenty-eight billion dollars in revenue. And the market erased a hundred and forty billion in a single day. Both of those things are true.

This is Tesorb Signal and I’m Lena Ruiz. Three things today. The record quarter and what it actually delivered. The miss that buried it. And why the market is done paying on faith.

Start with the headline, because it earned its moment. Revenue hit twenty-eight point two billion, up twenty-six percent from a year ago, blowing past what Wall Street expected. For the first time in its history, Tesla has now earned more than a hundred billion dollars over the last twelve months. We covered the delivery beat when it happened earlier this summer. The cars showed up.

Before the earnings call, we told you those deliveries were only the opening argument. The financial results were going to be the cross-examination. That cross-examination just happened.

Adjusted earnings per share came in at thirty-three cents. The street expected around fifty-three. That’s roughly a thirty-nine percent miss on the number investors care about most. Revenue hit a record. Profit didn’t come close.

Now, to be fair. Last quarter had roughly two hundred and thirty million dollars in one-time benefits, warranty reversals and tariff-related items, that didn’t repeat this time. Strip those out and the car margins were roughly flat, not falling. Tesla’s manufacturing cost discipline is still real.

And the spending wasn’t a surprise. Tesla told investors it would pour more than twenty-five billion dollars into capital expenditures this year. That guidance came months ago. Think of it like remodeling a house. You knew the kitchen was getting torn out. You budgeted for it. The bills arriving on schedule shouldn’t shock anyone.

Negative free cash flow of about a billion dollars sounds alarming until you look at where the money went. Capital spending more than doubled, up a hundred and forty-two percent. That’s factories, AI training clusters, and robotaxi infrastructure. This was the plan.

And here’s the context the selloff missed. Tesla’s stock was already limping into this report. Down for the year while the broader market climbed. The day after earnings, it hit its lowest close in nearly a year. The selloff wasn’t a sudden reversal. It was an acceleration of a slide that started months ago.

So why did the market react like the house was on fire? Look at the operating expenses. They grew forty-seven percent, nearly double the rate of revenue growth. When a company’s costs grow twice as fast as its sales, Wall Street notices.

Then there were the regulatory credits. A year ago, those credits contributed roughly two percentage points of margin. This quarter, they were worth about half a point. A hundred and forty-six million dollars. The federal EV credit expired last September. Fuel-economy penalties that other automakers used to pay Tesla went to zero. Those cushions are gone.

Put those two things together and you get a quarter with no financial makeup on. Revenue at a record. Costs growing faster than revenue. And the subsidies that used to paper over the gap, vanished.

This is the first time in years that Tesla’s income statement has been this exposed. Every dollar of profit had to come from actually building and selling things.

And the number that tells that story in one breath. Operating margin, one point four percent. A year ago it was four point one. That means Tesla kept about a penny and a half of every record-revenue dollar. A penny and a half.

On the earnings call, executives talked robotaxis in seven metro areas and confirmed that Cybercab production has started in Texas. But they offered no fleet numbers, no production counts, no timeline for when those services generate meaningful revenue. The future was long on adjectives and short on arithmetic.

One more number worth noting. Tesla’s adjusted results excluded a billion-dollar unrealized paper gain from its SpaceX investment. If you’re a bull, that money exists and it matters. If you’re a skeptic, it’s a reminder that one of the brightest lines in Tesla’s financial story came from a different company entirely.

And there are bright lines. The services business hit an all-time record margin. Energy storage is scaling. FSD subscriptions grew more than fifty percent. The pieces of the next Tesla are real. They’re just not big enough yet to carry a quarter on their own.

That’s the shift this quarter exposed. For years, the market gave Tesla credit for where it was going. Growth covered the gaps. Credits softened the margins. Investors paid for the story, not the spreadsheet. This quarter, the story showed up and the spreadsheet didn’t follow. The market stopped paying on faith.

So don’t argue with the deliveries. Argue with the margin.

Tesla right now is two companies wearing the same badge. One builds and sells more cars than it ever has, earns a hundred billion a year, and is scaling energy and AI at a pace that would make any startup jealous. The other keeps a penny and a half out of every dollar and has no timeline for when robotaxis or robots turn into real revenue.

Which company the market prices on any given day depends on one thing. The margin. When that penny and a half starts climbing back toward four, or five, or beyond, the record revenue becomes the headline again. Until then, it’s just a big number with a very thin bottom line.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

Got a tip or feedback? Send it to [email protected], we’d love to hear from you. This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. See you on the next one.

On a stretch of Gulf coast at the southern tip of Texas, nearly everyone works for the same company, lives on land it owns, and votes for leaders who draw a paycheck from it. Its name is Starbase, and a year ago it became a real, legal city. The vote was 212 to 6. This episode walks the lopsided vote, the very old American idea it revives, and the single power the whole thing was really about: who gets to close the beach.

On a stretch of Gulf coast at the southern tip of Texas, there’s a place where nearly everyone works for the same company, lives on land that company owns, and votes for leaders who draw a paycheck from that company. It has a mayor, a budget, and a spot on the map. Its name is Starbase. And a little over a year ago, it became a real, legal city.

This is Tesorb Signal and this is Lena Ruiz. Today, the town SpaceX built, one year in. We’ll walk the lopsided vote that created it, the very old American idea it revives, and the single power the whole thing was really about. It’s not the rockets. It’s who gets to close the beach.

Start with the vote, because the number tells the story. In May of last year, the few hundred people living around the SpaceX launch site went to the polls to decide whether to become their own city. The tally was two hundred and twelve in favor. Six against. Two-twelve to six. That’s the hero fact of this episode, and it’s worth sitting with.

Because that isn’t a normal election result. That’s a near-unanimous vote, in a place where most of the voters are SpaceX employees, on a question that SpaceX wanted answered yes. Nobody rigged anything. It’s simpler than that. When almost everyone in town works for one company, the company doesn’t need to twist arms. The town already agrees.

So Starbase became an official Texas city. Small, under two square miles, fewer than five hundred residents, most of them workers and their families. Its mayor is a SpaceX vice president, who ran unopposed. Its commissioners are SpaceX employees too. The people who run the city and the people who run the rocket factory are, quite literally, the same people.

Now, if this feels new and a little dystopian, here’s the surprise. It’s actually very old. America used to be full of company towns. Pullman, which built railroad cars. Hershey, which made chocolate. Whole communities where one firm owned the houses, the stores, and the streets, and employed nearly everyone in them. Starbase is a modern remake of a century-old model.

Let’s be fair about the practical case. SpaceX made a reasonable argument. It already ran the roads, the utilities, even some of the schooling and medical care out there. Becoming a city just formalized what existed. And a launch site coordinates closures and construction constantly. Doing that as a local government beats begging the county each time.

And to be fair to the residents, they wanted this. These are people who moved to a remote spit of land to build rockets, and they’d rather govern their own community than answer to officials an hour away who don’t share their mission. That’s not sinister. That’s a group of people voting for local control over the place they live. On its face, it’s democracy working.

But look closer at what that local control unlocks, and you find the real prize. For years, every time SpaceX wanted to launch, it had to close a public beach and a state park nearby. And the power to close that beach belonged to the county. The county sometimes pushed back. That was friction. That was delay.

Here’s what incorporation quietly changed. Alongside the city vote, Texas shifted weekday authority over closing that beach away from the county and toward the new town, the town SpaceX runs. The county kept the weekends. So on the days launches actually happen, the company that wants the beach closed controls the decision to close it. The referee and the player became the same person.

That’s the tell. This was never really about potholes or streetlights. It was about a public beach, and who holds the key to it. A rocket company that needs to close public land dozens of times a year just arranged things so that closing it is now a local matter, decided by locals who work for the rocket company.

We talked in an earlier episode about how, with Musk’s companies, the real story is often the real estate, not the product on stage. Starbase is that lesson in its purest form. The rockets get the headlines. But the quiet, durable win was jurisdictional. It was turning a patch of contested public coastline into private-friendly home turf.

To be clear, plenty of people are fine with this. The launches bring jobs and pride to a poor corner of Texas. But some longtime residents and environmental groups aren’t fine with it at all. They’ve watched a company gain formal power over beaches that used to belong to everyone. They took the fight to the state’s highest court. In June, that court sided unanimously with SpaceX. Both of those things are true at once.

So this is what a company town really is, once you strip away the novelty. It’s not a perk for employees. It’s a jurisdiction. It’s the power to make rules, close roads, and shut a beach, held by a company instead of a neutral government. Starbase didn’t just give SpaceX a home. It gave SpaceX a seat on the other side of the table.

Don’t watch the launch pad. Watch the city limits. Because the most important thing SpaceX built in Texas last year wasn’t a rocket or a factory. It was a border, and the authority that comes with it. And that kind of win doesn’t slip on a schedule or explode on a test flight. It just quietly holds.

That’s the signal. I’m Lena Ruiz with the Tesorb Signal podcast. For more news about Tesla, SpaceX, and Elon Musk’s companies, visit our website at tesorb.com.

Got a tip or feedback? Send it to [email protected], we’d love to hear from you. This podcast was developed with the help of using AI assistants, including the voice, and undergoes a detailed review during production. See you on the next one.