Notes:
Neil: This is one of my favorite podcast episodes of all time. Elon Musk embodies The Stockdale Paradox throughout this interview, confronting the brutal facts facing his missions while keeping absolute faith that he will eventually succeed. I’ve highlighted my favorite segments below.
- Book Recommendation: Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future by Ashlee Vance (2015)
- Link to Audio: iTunes
- Full Transcript: Vox
- This post is part of my series of Notes collecting quotes from my favorite documentaries, interviews, and speeches.
Quotes:
Kara Swisher: “What happens? What happens with you and Twitter?“
Elon Musk: “Well, I tweet interesting things pretty much as they come to me, and probably with not much of a filter. I look at it as a way to learn things, kinda stay in touch with what’s happening. It feels like dipping into the flow of consciousness of society.“
KS: “You pick fights with the press over Twitter, and then you have all your fans, of which there are many. Are you aware of what they do once you start them off?”
EM: “Well, I have to say, my regard for the press has dropped quite dramatically.”
KS: “Explain that, please.”
EM: “The amount of untruthful stuff that is written is unbelievable. Take that Wall Street Journal front-page article about like, “The FBI is closing in.” That is utterly false. That’s absurd. To print such a falsehood on the front page of a major newspaper is outrageous. Like, why are they even journalists?”
KS: What about what Donald Trump does, about “enemy of the people”? Do you look at it that way?
EM: “No.”
KS: “Just that you don’t like falsehoods.”
EM: “Yeah. There are good journalists and there are bad ones, and unfortunately the feedback loop for good versus bad is inverted, so the more salacious that an article is, the more salacious the headline is, the more clicks it’s gonna get. Then somebody is not a journalist, they are an ad salesman. Not a journalist, an ad salesman.”
KS: “What about things that are just critical of you that you don’t like? Do you think you’re particularly sensitive?“
EM: “No. Of course not. Count how many negative articles there are and how many I respond to. One percent, maybe. But the common rebuttal of journalists is, “Oh. My article’s fine. He’s just thin-skinned.” No, your article is false and you don’t want to admit it.“
KS: “Do you take criticism to heart correctly?”
EM: “Yes.”
KS: “Give me an example of something if you could.”
EM: “How do you think rockets get to orbit?“
KS: “That’s a fair point.”
EM: “Not easily. Physics is very demanding. If you get it wrong, the rocket will blow up. Cars are very demanding. If you get it wrong, a car won’t work. Truth in engineering and science is extremely important.“
KS: “Let’s talk about this year. What has gone on this year with you?“
EM: “It’s a very roller-coaster year.“
KS: “Why is that?”
EM: “We had the Model 3 production ramp, which was excruciatingly difficult. It is incredibly difficult to survive as a car company. Incredibly difficult. People have no idea how much pain people at Tesla went through, including myself. It was excruciating.”
KS: “And?”
EM: “Excruciating.”
KS: “And talk about that toll.”
EM: “Pretty sure I burnt out a bunch of neurons during this process. Running both SpaceX and Tesla is an incredibly difficult … You realize we’re fighting the incredibly competitive car companies.”
EM: “And the history of car companies in America is terrible. The only ones that haven’t gone bankrupt are Tesla and Ford. That’s it. Everyone else has gone bankrupt.”
KS: “What I want to get at is why you’re doing that. It’s not a trivial … Why do you think you want to push yourself that hard?”
EM: “Well, the other option would have been, Tesla dies.
KS: “Right.”
EM: “Yeah. Tesla cannot die. Tesla is incredibly important for the future of sustainable transport and energy generation. The fundamental purpose, the fundamental good that Tesla provides is accelerating the advent of sustainable transport and energy production.”
EM: “The success of Tesla is, by far, the biggest forcing function for the other car makers to get into —
KS: “100 percent.”
EM: “Yeah. Into electric cars. They’ve said so.”
KS: “You’re doing this to yourself because you think that the world depends … Not the fate of the world. You’re not a cartoon character.”
EM: “Without Tesla, this would still happen. There would still be a transition to sustainable energy, but it would take much longer. History will judge this, obviously, but I would say on the order of 10 years, maybe 20 years.“
KS: “So, pushing it forward by that much.”
EM: “Yes. I think it’s probably fair to say that Tesla has advanced sustainable energy by at least five years, conservatively, and maybe closer to 10, and then if we continue to make progress, we might advance it by 20 years. This could be all the difference in the world.“
KS: “What is the toll on you? What has been the toll on you and your employees? How do you think about that?”
EM: “It’s been terrible. This year felt like five years of aging, frankly. The worst year of my entire career. Insanely painful.“
EM: “For this past year, it’s been because of the Model 3 production ramp. Myself and others at Tesla, we had to go in and fix the mistakes in the Model 3 production system, and there were a lot of them. I personally solved a bunch. Jerome [Guillen] solved a bunch. Everyone helped, the entire team. Javier [Verdura], Franz [von Holzhausen], Deepak [Ahuja], everyone. It was … like, we had the legal team delivering cars in Q3. Todd [Maron] is great. There was a lot of people … Everyone had to basically go hardcore to solve the ramp.“
EM: “The amount of money that you lose if you don’t solve the ramp is mind-boggling, because you get hit from both sides. Let’s say you’re selling software or something, you don’t…”
KS: “That would be easier, Elon.”
EM: “It would be a lot easier. But in software you don’t have a bunch of parts. You don’t have a supply chain. If you don’t sell software, you lose the revenue, but you don’t incur a massive amount of cost associated with producing the software, ’cause you just make copies pretty easily. But for any large, complex, manufactured item, you have an entire supply chain, and that supply chain has on the order of six months of inertia. You place the order for parts for the car, including if you’re all the way with like two or three or four suppliers … We have on the order of 10,000 suppliers. It’s a crazy number. “We have to place the orders for how many cars we think we’re gonna build roughly six months in advance, six months before final assembly, ’cause if you go all the way down the supply chain… Then if you don’t actually make those cars, you still have all the costs. It’s like a flotilla of supertankers. The inertia of that is incredible. If anything happens to stall out the production progress, and that could be any one of those 10,000 suppliers, or on the order of 10,000 internal processes, if any of those is slow or wrong or whatever, you can’t make cars. You only need one missing thing, and whatever the slowest, least lucky part of the production process is, that sets your rate.“
KS: “But when you’re thinking about doing this incredibly complex thing, do you regret some of the things you’ve done to slow it down itself? Or was that unavoidable from your perspective? You know, some of your tweets. You attract attention. You really truly do, and some of it is self-inflicted. Do you not see it that way?”
EM: “Yeah, there’s no question there’s, like, self-inflicted wounds. In fact, my brother said, “Look, if you do a self-inflicted wound, can you at least not twist the knife afterwards?” You stabbed yourself in the leg. You don’t really need to twist it in your leg.”
KS: “So why do you do that?
EM: “It’s not intentional.”
KS: “Well, okay.”
EM: “Sometimes you’re just under a lot of pressure, and —
KS: “Your brother is wise.”
EM: “You’re not getting much sleep, you’re under massive pressure, and you make mistakes.“
KS: “Is that over? Do you feel like that’s over? Do you feel calmer now?
EM: (Sarcastically) “It’s totally over. I will never make another mistake again.“
KS: “No, I’m teasing you. But how do you … You look well. You don’t look under a lot of pressure. You seem rested.”
EM: “Yeah. Things are back to a hard work schedule, but not an insane work schedule. I was, there were times when, some weeks … I don’t know. I haven’t counted exactly, but I would just sort of sleep for a few hours, work, sleep for a few hours, work, seven days a week. Some of those days must have been 120 hours, or something nutty. You’re gonna go a little bonkers if you work 120 hours a week. Now we’re down to 80 or 90. It’s pretty manageable.“
“EM: “Yeah. 80 is pretty sustainable.”
KS: “Sustainable for you.”
EM: “Yeah. The pain level for hours increases exponentially. It’s like nonlinear above 80.“
EM: “I think at Tesla we’re doing pretty well right now. Tesla’s not staring death in the face. We’re in, I think, a pretty good position. We don’t want to be complacent, but it’s not … Up until around September, we were really faced with, like, “We must solve this or we’re gonna die,” constantly. I feel like we’re no longer in the staring-death-in-the-face situation.”
KS: “Talk about the new navigation feature.”
EM: “Drive on Navigation?”
KS: “Right.”
EM: “That’s I think one of the first major steps toward full self-driving. You can enter in an address, and from highway on-ramp to highway off-ramp, the car will change lanes. It will go from one highway to the next automatically and take the off-ramp automatically. It’s pretty wild. It’ll overtake a slow car. It’s basically integrating navigation with the Autopilot capability. That’s why we call it Navigate on Autopilot or Drive on Nav.“
KS: “What are the challenges that you face with these technologies now, from your perspective?“
EM: “Well, the main challenge has been improving the neural net so that we can recognize all types of objects from all eight cameras. There are eight cameras: Three forward, two on each side, and one rear. We’re running essentially eight neural nets of varying complexity. We’ve got to integrate the output of the neural nets into path-planning and then hook in the navigations, say, “Where do you need to go?” The big challenge has been solving a wide range of corner cases.“
EM: “Sometimes tar seams look like a line. Sometimes the lines are just painted wrong, for some reason. One of our biggest challenges, actually, with Drive on Navigation was dealing with forks and gores, where if a lane is splitting, you need to be confident that you’re going either left or right, not down the center. And the car will come to a halt at the first intersection. Now we’re integrating stop signs, traffic lights, being able to do, say, hard right turns or hairpin bends and that kind of thing.”
EM: “The car needs to drive better than a human driver using the same inputs as a human driver. Eyes are basically just cameras. All creatures on Earth navigate with cameras. A fish eagle can see a fish from far away and take into account the refractive index of the water, dive down and get the fish from far away.”
EM: “There’s no question that image-recognition neural nets and cameras, you can be superhuman at driving with just cameras.“
KS: “You don’t need anything else, from the government or from infrastructure or anything? I was recently talking to the Mercedes people. They were talking about sensors in the roads.“
EM: “Yeah, that’s hopeless. That would, at best, be a specialized solution, and whatever city puts stuff in roads … You can always make something work for a specific solution, like some special-case solution in some town, you can make that easy, but what you really want is a general solution for self-driving that works worldwide.”
KS: “That works worldwide. How do you look at the regulatory environment, because that’s another thing you have to be dealing with? Here in the U.S., in China, wherever else.“
EM: “I think the key thing for convincing regulators that self-driving is OK is to show billions of miles with a much better statistical significance in safety than human drivers. If the probability of injury is half that of the average human driver, then I think, probably, regulators will support self-driving… But it has to be a very large population. The statistical population of miles has to be very big, like billions, in essentially almost every possible case.”
EM: “I don’t really think that much about competitors. I just say like, you know, how do we make our cars as good as possible? How do we make sure we have like the best engineering and manufacturing talent in the world? You know, Tesla doesn’t do any advertising, or we don’t do any paid endorsements, we don’t sort of haggle for cars or anything like that. So we’re really reliant on the quality of product to sell. And I think it makes sense to sort of put our budget into advancing the technology to make the best possible cars.“
EM: “I’m not sure looking at competitors really helps. It’s sort of like the old adage with, you know, running, you don’t wanna … If you start looking at the other runners, it’s not good, you know.
KS: “Right.”
EM: “Like you can lose races because of that.”
KS: “Do you think about them? Do you think about them at all? Like Ford or Mercedes or anyone? Or if they’re doing anything that’s interesting, or Google? Which one of them, do you think, is the furthest ahead or closest to you all?“
EM: “I mean self-driving, maybe Google, Waymo? I don’t think anyone is close to Tesla in terms of achieving a general solution“
EM: “You really have to have a generalized solution. And best to my knowledge, no one has a good generalized solution except … and I think no one is likely to achieve a generalized solution to self-driving before Tesla. I could be surprised, but…”
EM: “The other car companies … I don’t wanna sound overconfident, but I would be very surprised if any of the car companies exceeded Tesla in self-driving, in getting to full self-driving. You know, I think we’ll get to full self-driving next year. As a generalized solution, I think.”
KS: “Will there ever be a serious competitor, from your perspective?”
EM: “In self-driving, I don’t think so. They’re just not good at software. And this is a software problem… It’s a harder problem, too, on the compute side. But they’re also not doing anything on the compute side.
KS: “Right.”
EM: “So it’s like … You do need an advanced sort of AI computer that’s very good at doing matrix multiplication with localized memory. So that —”
KS: “So they’re missing elements? What you’re saying, the generalized solution?”
EM: “Yeah, I mean … you need a specialized inference engine. Like the Tesla hardware 3 Autopilot computer, that will start rolling into production early next year, is 10 times better than the next best system out there at the same price, volume and power consumption. And it’s really because it’s got a dedicated neural net chip. Which basically, it sounds complicated, but it’s really like a matrix multiplier with a local memory.”
KS: “So the challenge you face is financial, though. Getting funding and stuff like that. And you’ve gotten … Saudis had bought a big bunch of your stock”
KS “But where do you get the money? Talk about the finances of doing this, because that’s what could really hurt you is not having enough capital.“
EM: “You know … I mean, as I said earlier this year, I think we will be cash-flow positive for all quarters going forward.“
KS: “All quarters going forward. So do you need more investment?”
EM: “No.“
KS: “Not at all.“
EM: “I don’t think so.“
KS: “Do you need to go private? Are you still contemplating that?”
EM: “We don’t need to go private, I think I …. going private would … I think we could execute better if we were private.“
EM: “the issue is that there’s a group of people who are quite smart, very mean, and have a strong financial interest in Tesla’s downfall. And what that results in is a constant attack on the Tesla brand, on me personally, on the executive team, on our cars. You know, every mistake we make is amplified. And this is not good. So the …”
EM: “Going private would definitely result in some short-term drama, but if we can avoid the distraction of … If we can avoid the brand damage of … Let’s say we’re private, and then we went public five years from now. Then the area under the curve of brand damage by short-sellers would be probably less than the short-term difficulty of going private in the first place.“
EM: “And then also being public, particularly when everyone at the company’s a shareholder, causes a lot of distraction when the share price moves around a lot.“
KS: “Do you have a motorcycle?”
EM: “No. I rode motorbikes a lot when I was a kid. So I did, like, dirt biking and then rode a motorcycle on the road. And then I almost got killed when I was 17, so … You know, most people are paralyzed, but depending on how you count it, the probability of death in a motorcycle versus —”
KS: “It’s quite high.”
EM: “It’s 25 times higher.”
KS: “All right. Well let’s get to rockets then. SpaceX. Last time we talked, you said you wanted to die on Mars, just not on landing. Which was a very funny joke, although it’s probably not a joke, it’s probably —”
EM: “Well, it’d be ironic if that had happened.“
KS: “Well, you know.”
EM: “Better not … I think we just be careful … I have to be careful about tempting fate, because I think often the most ironic outcome is the most probable.“
EM: “there’s a friend of mine, Jonah Nolan, who had this like modification of Occam’s razor where he said he thinks“the most ironic outcome is the most likely. And then also I think sometimes the most entertaining outcome is the most likely. Whether that entertainment is in the nature of drama, comedy or something else —”
EM: “Yeah, so this year’s been great for SpaceX.
KS: “Right.”
EM: “We successfully launched the Falcon Heavy rocket, which is the most powerful rocket in the world by a factor of two. So that’s twice the power, twice the thrust of the next biggest rocket. And we actually launched a Tesla — my Tesla Roadster to Mars orbit.”
EM: “I’m trying to listen to the reality simulation, which is do the most entertaining thing.“
EM: “And you know, we had a lot of, you know, like playing David Bowie. And where, like, “Hitchhiker’s Guide to the Galaxy” stuff in there, we’ve got the Asimov Foundation series etched in glass. Like there’s a lot of like little things.”
EM: “The … like the reason we did that is actually because, normally, when a new rocket is launched, you just put a dummy payload, which is like a block of concrete or something. Like, literally, I think when Boeing launched the Delta IV Heavy, they launched a block of concrete or something… So we were like, okay, what is the least boring thing we can launch?“
EM: “And then next year, the exciting things are we’re gonna be launching astronauts for the first time to the space station.“
EM: “the United States has relied upon the Russian Soyuz, which actually recently has had some issues.“
EM: “for the first time since the Space Shuttle, a U.S. vehicle will transport U.S. astronauts to orbit.“
KS: “What do you think of the Space Force? The Trump Space Force?”
EM: “Well, this may be a little controversial, but I actually like the idea. I think it’s cool. You know, like, when the Air Force was formed, there was a lot of like pooh-poohing, and like, “Oh, how silly to have an Air Force!” You know, because the aircraft in World War II were managed by the Army…”
EM: “people today may not realize back then it was wildly panned as a ridiculous thing to create the Air Force, but now everyone’s like, “Obviously you should have an Air Force.” And I think it’s gonna become obvious that we should have a Space Force, too.
EM: “it’s basically defense in space. And then I think also it could be pretty helpful for maybe expanding our civilization … You know, expanding things beyond Earth.”
EM: “You know, the Space Force could be something that … Like, I think we could just have a base on the moon, for example. A base on Mars. Be great to expand on the idea of a Space Force.”
EM: “Like, you know, I think … for explorers … anyone who has an exploratory spirit, and I think that especially applies to a country like the United States, where you know it’s kind of the distillation of the spirit of human exploration. I think the idea of being out there among the stars and among the planets is very exciting.“
KS: “Do you think the Trump administration has a commitment to it, or is it just, “Let’s have a Space Force?””
EM: “I don’t know … I haven’t had detailed conversations.”
KS: “Right. Right.”
EM: “But I do think it will become obvious over time that a Space Force is a sensible thing to do.”
KS: “And, Mars. Last time we talked it was 2024, was it? That you talked about getting there?
EM: “Yeah, we’re still aiming for 2024.“
EM: “I don’t know if I will go or not. It may be just an unmanned mission, you know. I’m not sure if there’ll be people on board or not. But there is a Mars rendezvous opportunity, ’cause you can only do a launch to Mars roughly every two years. So around the 2024 timeframe, there’s a rendezvous opportunity for Mars, which hopefully we can catch. There’s one in 2022 —”
EM: “Well, I suspect by the time we make enough progress to wanna try to do it, I suspect that the U.S. government will be interested in doing this, too, or being part of the effort. But I think … the vehicle that we’re designing right now, which is sort of code-named BFR, I’m thinking of changing the official name to Heart of Gold.“
KS: “Okay. As in Neil Young?”
EM: “Well, I do like that song, but it’s more like the spaceship in “Hitchhiker’s Guide to the Galaxy.”“
KS: “What was BFR? What did BFR stand [for]?”
EM: “BFR is, to some degree, a Rorschach test on acronyms. But officially, it is the Big Falcon Rocket.”
KS “How do you look at competitors, say, like what Jeff Bezos is doing?”
EM: “Yeah, I certainly think NASA should continue to exist, NASA does a lot of really useful things, and these go beyond astronaut transport. There are missions to rovers on Mars that are thanks to NASA. There are these planetary probes, there’s the Hubble Telescope, there’s a … So, NASA does a tremendous amount of good.“
EM: “I think it’s high time that we went beyond Earth orbit again. I think it’s very exciting and inspiring, and I think it really gets the whole world fired up.“
EM: “I think it’s great that Jeff is spending lots of money on space. And they are a competitor, but it’s good that he’s spending money on … that he’s spending a lot of money on developing rockets. I think it will encounter some challenges getting to orbit, it’s remarkably difficult getting to orbit. But he has the resources to overcome those difficulties.”
KS: “You’re not buying a newspaper, are you?
EM: “No, I don’t generally acquire things. I create companies, but I don’t really acquire them. So I wouldn’t … I have no plans. It does seem to be popular these days.”
EM: “I think this is really the key to getting around the city very fast. You’ve got to go 3-D. Essentially, we have a 3-D … our offices are 3-D and dense, but we then have a 2-D road transport network, so it’s … And everyone would just go in and out of the buildings at the same time, so, naturally, you’re going to have traffic.“
KS: “Right, like a subway system?”
EM: “Yeah, but even subways tend to be essentially two-dimensional. You’ll have a subway cross another subway, but they’ve never really tried to make many layers of subways. The cost of tunneling, historically, has been prohibitive. And they’ve also been incredibly slow. So, if it takes … I don’t know how long. Say the New York Subway had a one-mile extension or something, and it cost —”
EM: “So the typical cost for a subway, per-mile cost for a subway in the U.S. has been about a billion dollars a mile, so that is not a very scalable solution.
EM: “You can go further down than you can go up. So the deepest mines are much deeper than the tallest buildings. But, really, the key is a massive improvement in tunneling technology. That’s the linchpin, that’s fundamentally what it amounts to.”
EM: “I always look at things from a fundamentals-of-physics standpoint, and so if you sort of apply physics’ first principles to any given technology endeavor, you can sort of envelop the possibilities.“
EM: “the first question I’d ask is, “Well, is your tunneling machine power-limited or thermally limited?” This was a very obvious question from a physics standpoint. Nobody knew.”
EM: “Yeah, essentially taking rocket technology and automotive technology and applying it to drilling technology.“
KS: “Right, I think of that a lot about house construction, how slow it is. Why is house construction so slow?”
EM: “Well, you can make house construction crazy-fast if you’re in a factory.”
EM: “Yeah, construction in general is … I think there’s a lot of potential for disruption and for entrepreneurs to enter construction in general. There’s a tremendous amount of opportunity.”
KS: “At what cost?”
EM: “I don’t know, I think it’s probably … Excluding the equipment, probably cost us $10 million for a mile.“
EM: “OpenAI was about the democratization of AI power. So that’s why OpenAI was created as a nonprofit foundation, to ensure that AI power … or to reduce the probability that AI power would be monopolized.“
EM: “There is a very strong concentration of AI power, and especially at Google/DeepMind. And I have very high regard for Larry Page and Demis Hassabis, but I do think that there’s value to some independent oversight.”
KS: “Okay, in this reality, when you look at it, how are you feeling about the future, when what appears to be reality …?”
EM: “For some reason, I feel optimistic. And I’m not sure if that is irrational or not, but that is my —”
KS: “Why are you feeling that?”
EM: “My current gut feel is weirdly optimistic.“