Aug 1, 2017 · 13 minutes

Last week, Tesla released its long awaited Model 3.

It was an important milestone for a company that started as a luxury product and has long hoped to steadily bring down the price of its cars and make them more mass market. While most major automakers have also jumped on the electric bandwagon-- some using Tesla’s technology-- even Elon Musk couldn’t have anticipated the 20+ company slugfest to dominate a self-driving car era. There are hundreds of billions in collective market cap and paper net worth riding on who wins that.

I don’t invest in private or public tech stocks, but my metaphorical money would be on Musk, even though he recently admitted his company was overvalued based on the current business.


Listening back to my interview with Musk and other Valley leaders for our recently launched Beyond The Series B podcast series, I had a number of realizations.  One was how quickly the financial landscape changed.

In 2012, our guest was [Pando investor] Marc Andreessen and I asked him about a debate he’d recently had with Peter Thiel about whether the Valley was doing enough innovative stuff, or just merely tweaking new ways to share photos.

Andreessen’s answer was more illuminating in hindsight than it even was at the time:

When I debate with Peter, he says, "We want to fly in cars, we got 140 characters," so therefore I can't resist. I attack flying cars and I defend Twitter, which grates him up to no end.

In a sense, I very much disagree with what he says, and then in a sense, I very much agree with it. What I disagree with is I think the innovation that's happening today in Silicon Valley is very important, profound, deep, and important. I'm completely unapologetic about the idea.

For example, Uber and Lyft. Lyft is the company we work with, we're very excited about. I think these companies are transformational for transportation. I think they're fun. We can talk about that for a long time. I think Airbnb is transformational for real estate. I think Bitcoin, transformational for financial services.

These companies have this incredibly nice property that when you can apply software, you can do it in a very cost effective way. Software's not very expensive.

The capital efficiency of having a small group of software programmers that build amazing software, who then go in and do something in an industry that's 100 billion or trillion‑dollar industry, existing venture capital structure and framework is very good at doing that.

It works very well when it works, and I think that's very valuable. I always accuse Peter of dismissing all that stuff out of hand, which is probably an overstatement.

The part that I struggle with, and I'm on the verge of agreeing with, it's SpaceX, it's Tesla. The trenchant critique comes from Larry Page, Elon, and Peter. It's like, "OK, software. Got it. What about electric cars? What about the Hyperloop? What about the SpaceX private rocketry?"

What about these bigger, more transformational things? In particular, what about the things that operate more in the real world? What about the things that are really going to affect natural resources, pollution, livability of cities, and all the things that are outside of whatever's running on the screen?

I think that there's a validity. Google is doing the self‑driving car. The self‑driving car is going to work. By the time it works, it will have cost hundreds of millions, and possibly billions of dollars to make work.

The self‑driving car is not a lean startup. Not in any way. It's entirely contained in Google because they have the money, and the staying power, and the patience to be able to fund it over years with a lot of money.

SpaceX and Tesla were not lean startups. They were very big, ambitious. They raised a lot of money. There's two problems. One is where do you think you can actually do this? That gets into things like where is it too regulated?

Those guys, all three of them, have more of an inclination to say, "Yeah, just power through the regulation. Just figure out a way. Give the world a better answer and let the world adjust," but how do you calibrate to that and not blow up an enormous amount of capital?

The other question is the venture capital funding model can fund really big companies. Facebook raised a very large amount of money as a private company, more than most people realize.

The problem with venture capital is the structure today for the bigger things is the assumption is every new round of capital has a new lead investor. The assumption is something has gone wrong if you don't have a new lead investor every step of the way. Raise 5, and then 10, and then 20, and then 40….

The big question, the question I'm noodling around, is what about the efforts where you have to say, "This thing is going to take $300 million?" It just is. There's no shortcut and there's no minimum viable product. It is going to take $300 million, and that $300 million has to be reserved ahead of time.

The good news is you know it's going to be there. The bad news is you've got to think hard about how that money gets staged and deployed in. You've got to think hard about what the milestones are. 

There are a couple of interesting things to parse here. One is that barriers to raise billions of dollars as a private company seemed insurmountable just five years ago. Andreessen is a man who’s frequently criticized of living too far in the future when it comes to technology. His firm has been accused of driving valuations up. His first major deal-- a buyout of Skype-- was hardly a typical venture capital deal. And yet, even he couldn’t imagine a scenario where $300 million in capital would be readily available for private companies. That should tell you we are in pretty untested waters here.

The other is that highly-regulated industries were seen as too risky, versus a positive in a wave of funding obsessed with “disrupting” regulation and up-ending stagnant trillion dollar industries.

But the most valuable part of Andreessen’s comments came next. He continued….

Those companies have to be run completely differently because the stakes are so much higher. What kind of entrepreneur can do that? Elon has proven he can do it. There are not a lot of other people in the Valley today who have done things at that level of scale.

There's a different kind of entrepreneur. There's a different kind of idea. There's a different kind of financing method. I think we'll all collectively figure it out, but we don't actually have it today...

...Venture capital as it's configured today is very good at picking up science that's been developed, largely with federal funding, for 30 or 40 years, and then doing the last 2 to 5 years of applied development to get it into a commercial product. We're really good at that.

The clean tech, the energy stuff largely not working illustrates we're not that good at picking something up that either has not had a lot of science funding behind it or is going to take 5 or 10 years to develop, as opposed to 2 to 5 years. The question is can we stretch the boundaries of the scope of the projects that we can fund, and support, and staff without losing our minds?

There's the other side to this, which is arrogance. We have what everybody thinks is a piece of modern art in our main conference room but it's actually a Solyndra solar panel that I bought off of eBay the week Solyndra went bankrupt for 10 bucks… They vaporized a billion and a half dollars, including taxpayer money, building cylindrical solar panels.

There is a point where you stretch this stuff too far. Somewhere between the next photo sharing app and Solyndra, somewhere in there, there's a sweet spot.

What I argue with Peter about, or I argue with people who say things like...It's not sufficient to say, "This is missing." We have to figure out how to actually do it. We don't yet actually have the formula yet for how to do that.

What you get is you have a very special entrepreneur like Elon, and then he can do it. What I always say is, "OK, who is the next Elon? And then we'll talk." 

Here’s what jumped out at me five years later revisiting this interview: We only solved one part of this. The money. It’s incredibly common -- even easy-- for companies without proven economics or business models to raise hundreds of millions of dollars all of the sudden. And years after everyone expected the unicorn economy to crash, we just had the biggest quarter of mega deals over $100 million since 2015.

And Softbank’s new $93 billion fund seems to only be escalating this. Softbank recently invested $500m in WeWork, $2bn in Grab and is reported to be seeking a “multibillion dollar” stake in Uber.  

Put differently, there seems to be no end in sight for the trend of Silicon Valley bros suspending adolescence, having zero accountability, and still raising all the cash in the world. “Crushing it” is easy when no one ever has to see your numbers or post a return and there’s dwindling board accountability.

Note what hasn’t changed: The types of companies or the types of founders, the other two parts of Andreessen’s equation of what would give us flying cars.

He specifically cites Uber and Lyft as software problems, not the kinds of companies that Elon Musk is building, not the kinds of things that Thiel lamented was missing in the Valley.

And he’s right: Uber is a front end app at best. Core parts of the app like mapping and payments come from other companies. Lyft and Uber have regularly swiped features from one another. Both offer commodity products where the same drivers provide the same service for roughly the same price. While socially transformative (even destructive) companies, there is almost no hard technology problem they’ve solved. We’ve connected cabs to smart phones. And it cost VCs more than $12 billion in the US alone.

Andreessen talked of companies that would simply require $300 million to solve a dramatic technology problem, the kinds VCs so rarely fund. Instead those same VCs have invested more than $8 billion in something that certainly didn’t need that much capital, as seen by Uber’s hiring Beyonce for parties and burning billions of dollars a year to fail in overseas markets.

“We were promised flying cars and all we got was 140 characters” > “We spent $8bn + $4.5bn for Uber and WeWork and still don’t have flying cars.”

I’m not sure anyone has said it to me quite like Andreessen inadvertently did five years ago: We thought cash was the problem. But apparently not.

Meantime, something also jumped out at me from our interview with Elon Musk-- who Andreessen described as essentially the only founder on the radar capable of building these kinds of companies. (And presumably, worth these insane amounts of private capital.)

We were talking about Kleiner Perkins funding Fisker over Tesla, and how companies like Fisker and A123 were starting to struggle, while Musk’s companies were pulling away. I asked what he’d done differently. His answer was... illuminating, particularly taken along with Andreessen’s comments.

Bear in mind, as you read his words, Tesla is the only Silicon Valley company in the autonomous car race that is actually building cars. Andreessen gave Google credit in his quote, but even Google has punted on building the actual cars. You can basically substitute “Uber” for “Fisker” here...

Tesla's a really different company to Fisker. Tesla is a hardcore technology company, and we do really serious engineering. You only build value in a company if you're doing hard work to solve tough problems. That's why companies are valuable. It's why they should be valuable, and largely is why they're valuable.

BAM! I’ve heard privately from mutual friends that Musk has been furious that Uber is valued at more than Tesla. This quote-- stated years earlier-- helps explain why.

He continued...

We do real manufacturing as well. The Model S, coils of aluminum and plastic pellets come in, and cars come out. We do real, hardcore manufacturing, quite vertically integrated. We did all the vehicle engineering, all the powertrain engineering, all the software.

We also of course do the styling and the design of the car. We've got a great design studio. Our head of design, Franz von Holzhausen is awesome. In the case of Fisker, it was headed by Henrik Fisker. He's a designer, so he's good at the styling of the cars, but he thinks it's all about styling, and it's not.

The reason we don't have electric cars is not for lack of styling…

...In any industry, there are only a few companies that get big and succeed. It just seems that in the case of Tesla, we were focused on making great products and solving hard engineering problems.

As evidence in Tesla for the fact that we solve hard engineering problems is the fact that Toyota and Daimler ‑‑ Mercedes ‑‑ buy electric powertrains from us. If this was easy, they would just do it. 

This is one reason my eyebrow raised recently when Musk talked about the Boring Company-- his seemingly on a whim new startup that is tunneling through the earth to solve transportation problems. That and the fact that Musk first tweeted that he had gained approval to build a Hyperloop from New York to DC only for those cities and two others to contradict him.

But despite the maybe-maybe-not messaging, Musk still enjoys the benefit of the doubt from most commentators, given that he has actually publicly shared details of his Boring technology and - well - the fact that he has actually built electric cars and rocketships. The Hyperloop may still be vaporware but you’d be brave to place a bet on that, against Musk.




Musk first talked about the hyperloop at that same PandoMonthly, insisting-- as he would for years-- that he wouldn’t be the one to build it. I have to wonder if that Tweet is the beginning of a reversal. Musk is certainly not known to just solve part of the problem… the least technically challenging part, at that.

Last week, I wrote that the top ten private companies in the US are worth an astounding $199 billion. Unless you are insanely bullish on-- uh, WeWork at $20 billion-- this is potentially a larger ticking time bomb of evaporated paper wealth than all the other US unicorns combined. This puts the argument that Tesla-- which is surpassing legacy auto-makers in market cap-- is overvalued in perspective. Particularly since Tesla is a proven public company. I’ll put it this way: If you think Tesla is overvalued, you’d go nowhere close to an Uber IPO...

All of this is one reason I wrote last week that SpaceX was the only decacorn whose valuation seems to make any sense. It’s the only one that seems to match the playbook Andreessen spoke about, with the possible exception of Samumed. (If it works.)

There’s a reason Musk reportedly scoffed when Kalanick called him about teaming up to build a self-driving car to rival Apple. They are playing fundamentally different games, and Musk has seen this movie before.