Ten years ago, in December of 2009, I walked into the SpaceX headquarters wearing a suit and holding four copies of my resume. I was there interviewing for a job at Tesla; at the time their Los Angeles office was co-located inside SpaceX. The receptionist took me to the cafeteria just outside of mission control to wait for my interviewers. A guy wheeled a rocket engine by me, making a turn to avoid the FroYo machine and a full-size replica of an Iron Man suit signed by the cast of Iron Man 2. Mickey Rourke signed the crotch. I realized pretty quickly that Iron Man and I were the only two people in the building wearing anything resembling a suit.

I thought this company’s success was unlikely, but that’s part of what made it interesting. I was far from the only one in doubt; all of my friends in the automotive industry thought I was making a huge mistake. A colleague who had worked with Tesla through a software supplier told me, “They have no idea what they’re doing.” My OEM friends had variations of the same theme that could be summarized as: “Look, nobody is buying electric cars, and if they do, we’re just going to swoop in and crush you.”

This fine blog you’re reading right now called the Model S “Vaporware” and and . Mitt Romney, the Republican presidential candidate called Tesla a “” in a televised debate. Even Elon Musk said he thought Tesla had a chance of success.

Looking back on the past 10 years, it’s tempting to say that they were all wrong, but “wrong” doesn’t feel like the correct word. They weren’t wrong. Except for Romney, he was a loser. But everyone else was making a reasonable guess based on the data at the time. They were “wrong” only because something wildly improbable happened.

Tesla spent much of the beginning of the decade dangerously close to failure. It . Investing in Tesla just after its IPO was an ill-advised gamble, as Tesla had just one Model S prototype and an empty factory to go up against huge legacy automakers. None of the reasonable evidence at the time pointed to a $75 billion company. It was a gamble on a dream.

After many late nights and innumerable stokes of luck, people started buying cars and the company got some financial traction. It seemed like overnight that the valuation skyrocketed, new EV companies started showing up all over California, and legacy automakers started opening up offices in Silicon Valley. It felt like the fledgling California EV industry had gone from a standstill to full speed at a rate that seemed, well, ludicrous.

California has had an automotive industry almost since there has been an automotive industry, particularly as it played beachhead to Japanese companies setting up headquarters here in the 1950s as well as design offices and over a through the years. But this was different. This was California companies creating a new type of car business.

I was at Tesla for the first half of this decade and since then I’ve been bouncing around a California EV industry that didn’t even really exist at the beginning of it. From the outside, it might still look like a few startup companies struggling to make cars and to be profitable, but from the inside it is unrecognizable. Tesla made big waves this decade, and companies like Rivian, Lucid, Canoo, and others are working hard to amplify them.

At the dawn of mass production, America had dozens upon dozens of auto companies, very few of which survived. A similar thing has been happening in California over the last decade, and we’re already seeing attrition from the gold rush to build EVs and EV companies.

For a while, money was coming in from all over to fund new EV startups. Things got really interesting as an incestuous job market began to grow; it was a good idea not to piss anybody off because you were probably going to work with them again somewhere else.

Some startups were luring away engineers with tens of thousands of stock options. Many of those stock options are already worthless, and the surviving companies have learned that making cars is really, really hard. And very expensive.

Ten years is a long time, but it is still weird to me how easily people forget what they thought was impossible a decade ago. Those same OEM friends that had dismissed electric vehicles now talk about their companies going fully electric as though it had always been a foregone conclusion. On the other hand, it’s also surprising how often I hear the same things I heard back then: Tesla is and weeks away from , or that the vehicle and the company will never make it.

Tesla still isn’t consistently profitable despite multiple promises, and almost none of the other surviving startups have sold any cars. But thinking back to that office inside SpaceX with a vehicle engineering team of a dozen people and the most cartoony CAD of what would eventually become the Model S, it all seems so impossibly different: hundreds of thousands of cars sold, multiple major factories, 45,000 employees, a global charging network, and a market cap twice as big as Ford. One of the cars is in fucking space.

And it’s still weird to see a dirty Model S with body damage because I can’t help but think of it as a brand new car. It’s strange to hear big OEM car companies say they’re going to be going all-electric in the next few years, because for so long they teased EVs they had no intention of making, or begrudgingly churned out compliance cars.

I think for most people in the automotive industry these past 10 years have been unexpected. But, for the consumer, the industry hasn’t really changed that much. Walking around an auto show recently I couldn’t believe how many internal combustion vehicles there were. In that respect, a look at your local car dealership probably doesn’t look much different than it did 10 years ago. Almost every vehicle has a gas cap, a few of them are hybrids, but for many brands, there are still no electric options.

For some reason, everyone talked to me about things like they were two years away. When I started at Tesla, the Model S was two years away. Two years later the Model X was two years away. Then the Model 3, the lower-cost car that would actually change things was two years away. But people don’t buy sedans, they buy crossovers and the crossover was another two years away. Most of these two years ended up being three years (). Now it’s the legacy OEMs that are two years away. Mercedes is going to have by 2022, Jeep will by 2022, Volkswagen by then.

The auto industry is a big ship with a tiny rudder. It takes years to make a new car, and at least a decade to change the industry’s collective mind about anything substantial. The ship has been rocked by those California waves so much that it had to make a major course correction, one that will be much more obvious over the next decade.

All my years in the industry seem like they should give me some ability to predict its direction over the coming years. But looking back to 2009, I find it hard to fathom what might be two years away at the end of 2029.