Opening day of Air Force Space Pitch Day. The two-day event was hosted by the U.S. Air Force to demonstrate the Air Force’s willingness and ability to work with non-traditional startups. The “Fireside Chat” features Lt. Gen. John F. Thompson, Space and Missile Systems Center Commander, and Elon Musk, Space X Chief Engineer. The chat covers the future of space, space industry, how to find talent, and various other topics.
We would like to invite General Thompson and Elon Musk to the stage. Round of applause.
All right. Well, good morning, everybody. We got about 30 minutes with Elon here. I love the soundtrack that we’re listening to throughout the day. We’ve got the soundtrack to the Martian, the soundtrack to Interstellar, even the soundtrack to the Stranger Things. Hopefully that has lightened the mood a little, and hopefully y’all have had the opportunity to enjoy the expo, some of the public pitches and some of the panel sessions that have occurred throughout the day. Our guest today, which needs no introduction. In the Forbes Innovation List for 2019, he was co-number one on that list with Mr. Jeff Bezos. And if you’d care to comment on that later, you’ll have the opportunity. (audience laughs) But I will remain silent on that. He’s obviously the founder of a few companies you may all be familiar with, like Tesla Motors and SpaceX. He’s also spent time on the board of a non-profit, Open AI, and sponsored innovative competitions like the Hyperloop Initiative. Elon, thank you so much for being with us today.
Thanks for having me.
So, we’ll just dive right in unless anybody has any objections to that. I just have some questions, we’re gonna be talking about primarily innovation, but interspersed amongst some of our questions on innovation will be some questions on leadership. Some questions on culture, especially in the small business growing to a large business kind of environment. And then I myself never got these questions until five years ago, but I may throw you a couple work life balance questions, quality of life questions. (audience laughs) Without further ado. Elon, a critical question that all businesses face when tackling challenges during product development is do we invest the time and resources to do this work internally, or do we contract it out and have an external partner do it with us? Can you talk to the audience a little bit about what kind of decision process you go through, whether you’re deciding to do it in one of your companies or contract it out to somebody else?
In the beginning, we try to contract most things out. Both at Tesla and SpaceX, we started off with most things contracted out, and that didn’t succeed. We in-sourced more and more of the time. In space particularly, the supply chain is not great. ITAR restricts you to really working with US companies. It’s very difficult to work with non-US companies because space rockets are a weapons technology. You have a very limited set of suppliers. And to the degree that you have legacy parts, you inherit the legacy costs and limitations. That required us to in-source most of the rocket. There’s probably only less than 10% of the rocket is coming from the space supply chain at this point. The automotive supply chain is better. Consumer electronics is a lot better. Wherever there is a lot of competition, the supply chain is better. It really depends on the part. It’s important to bear in mind that legacy components bring legacy costs and limitations. If you want to have something revolutionary, you can’t do catalog engineering. Catalog engineering is not- has limited ingredients.
The legacy industrial base isn’t able to adapt fast enough to what you’re trying to do, so you in-source it. Are there any special techniques or cultural aspects of that in-sourcing that you’ve found to be particularly beneficial in Tesla or SpaceX?
I really think, if you want to advance technology, you’ve got to recruit the world’s best engineers and then create an environment which enables them to be as innovative as possible. The reward structure, it needs to really reward and encourage innovation, and punish lack of innovation. So it’s gotta be both. Stick and carrot. Ideally more carrot than stick. That was my new philosophy. (audience laughs) More carrot, less stick. (laughs) The thing is that in a lot of companies this is never-
Risk reward assymetry where bold moves, if they go wrong, are punished, but sort of keeping your head down is not punished. That’s not good. That will result in a very conservative outcome. It has to be a requirement that there’s significant advancement, and simply the lack of doing something significant is bad.
Big companies just become really conservative over time, and really the government is a company in the limit. The government is the biggest company.
We have some big companies in the audience, but we have a lot of small businesses in the audience.
So not part of the, if you will, the legacy supply chain for rockets or whatever other innovation you or a company you were associated with might be producing. Is there some advice or counsel that you could give to some of these small businesses about how to market to growing space partnerships like SpaceX or other companies that might be willing to admit them to the supply chain if they don’t have that old think or that old process mentality?
For SpaceX or Tesla, I mean, if somebody’s got a component that’s better than what we’re making internally we would love to buy that. For sure. We would just reallocate the resources that we’re working on that component to do something else. For example, on Falcon 9, we used to build the landing legs, the big carbon fiber landing legs that fold out, ourselves, and they were contracted out to All American Racing. They do like Norton racing, like racing cars.
They’re doing a great job. They’re doing better than us. We handed the building legs to them and worked on other things. This is by no means-
It would be completely insane for us to want to continue to make a part ourselves internally that is inferior to a part that is available externally, though it matters. We’d love to outsource more. That would be great.
So that’s a pretty high bar, but if you do it better than SpaceX, then they’d like to talk to ya. Okay? (audience laughs) Elon, given your extensive experience in the nonprofit arena, with companies like Open AI that I mentioned in the introduction, and your ability to tap in to the academic environment, that cutting edge innovation environment with things like the Hyperloop Competition, are there smart ways for the folks in the audience, this community of space professionals, to be engaging with different sources of innovation in the ecosystem? How do we better cultivate those sources of innovation so that we can take advantage of them in our nation’s space capabilities?
I wouldn’t say I’m like an expert in nonprofits, but Open AI was really intended to mitigate the risk of artificial general intelligence. I hope that that’s what it does. There’s some chance it may amplify the risk, but hopefully it diminishes the risk. I have a sort of foundation that gives away money, mostly anonymously. Though I wouldn’t say I’m an expert on this. What I’m obviously trying to do is figure out the set of actions that increase the probability that the future is good, and take those actions.
Okay. Are there techniques you personally use to identify sources of information. I mean, is it just reading or seeing what’s out there? Do you have continuous market research going on in your companies or in your private life about, hey what’s out there, that I should be sponsoring or taking advantage of to making things better than they are?
I do zero market research whatsoever.
We don’t even have a marketing department really. Or an advertising. It’s just like wouldn’t it be a great car or a great rocket. It’s like we try to think of like where’s the plutonic ideal of say the perfect rockets or car. What characteristics would it have, and then make that. I find that if you do that, people want to buy it.
That’s it. We’re gonna come out with the Tesla pickup truck, or we call it Cyber Truck. I mean it looks like an armored personnel carrier from the future. (audience applauds) Yeah, it won’t look like a normal truck. People might not like it. They might not. I like it.
Sounds like a couple other people do, too.
It’s gonna look like it came off a movie set. One of those on the road, you’re like whoa, what’s that thing? (audience laughs) It’s a little bulletproof.
Okay. (audience laughs) I look forward to seeing it, and perhaps driving it at some point in the future.
If I can elaborate on my-
A lot of times, I think people try to make products that they think others would love, but they don’t love them themselves. If you don’t love the product, you should not expect that others will. (audience applauds) You know your own heart, and if it’s compelling to you it will be compelling to others.
Okay. Great. From a leadership perspective, obviously you started as a leader of a small business. You grew a number of small businesses into large businesses. From your early days at Zip2 and PayPal, leading very very small teams of developers, sometimes doing a lot of the development yourself, hands-on, to today as you lead very large companies like Tesla Motors and SpaceX, has your leadership approach changed at all? Or is managing those small teams, do you try and keep that culture and your leadership style the same throughout?
Well, it definitely has to evolve as companies get bigger. Tesla’s around 45,000 people and SpaceX is almost 7,000 people. When a company’s little, then your skill as an individual engineer can make a very big difference. When a company is large, you have to kind of teach a lot of people to do it. You have to be a force multiplier as opposed to-
If you have a little band, like a dozen swordsmen or something, and you’re a swordsman, okay great. That’s going to make a difference in a little battle, but not if he has 12,000. You can’t like run along the entire battle line, it doesn’t work. So you have to try to teach people en masse different approaches, and make sure you reward the right behavior. The reward structure is incenting the right behavior. This is incredibly important. Economics 101 whatever you incent is likely to happen. In fact, it would be bizarre if it didn’t. Statistically speaking it will definitely be what happens. The incentive structure, reward and punishment, must be sensible. And, this sounds very obvious, but in most organizations the true reward and punishment is not sensible, and then people are puzzled why they do not get the outcome that they want. It’s because the reward structure isn’t right.
Okay. As businesses grow, as your businesses have grown, and it becomes less and less about the engineer talking to the program manager one on one in a very small team environment and it becomes conversations between different teams in your enterprise, much larger groups of people. In many organizations there’s a tendency that seams will develop, and leaders have to manage those seams or insist that their teams manage those seams. How have you managed the seams in the large businesses you manage currently?
I think that’s a great point that you’re making in the question. In any given product, you can see the mistake, the organizational errors manifest themselves in the errors in the product. I see this in our own products. For example, we’ve got like a top cover of the battery and we’ve got a bottom cover on the car. Okay, that’s an example of an organizational error. We should only have one cover. You don’t need a box in a box. But there’s the battery team and the body and chassis team, so they both made a cover. You’ll see sort of flanges and joints and various things that don’t make sense or where things are doubled up, and where you have substance optimization instead of system optimization. To counteract that, which is not easy, I actually insist that the teams step on each other’s toes. In the rocket, the propulsion team, the engine team has to go partway into the airframe, and the airframe team has to go partway into the engine. It’s hard to get people to do that. Essentially they have to basically offend other people in the company and not be offended themselves.
This is very important. One of the things I try to propagate is that everyone should be chief engineer, meaning everyone should have at least a cursory understanding of the whole rocket or the whole car, even though they may have deep expertise in one arena. Then they’ll be able to tell if they’re optimizing for the product as a whole. In general, I find you can summarize the key characteristics of each discipline-
You can simplify it down to a few principles. Richard Feynman used to say that you really know your subject if you can explain it to a smart 10-year-old. But if you try to disguise your expertise in obscure language, then maybe you don’t understand what you’re talking about. For example, when, say, the rocket’s coming back from orbit, you have center of pressure and center of mass, and it effectively-
It all sounds very complicated, but basically you just have a seesaw. (laughs) If the thing’s of approximately equal density, it’s just gonna be-
If you put mass on one side, it’s going to tip that way. If you put mass on the other side, it can tip the other way. Think of it just like a seesaw. Then we’ll see the mass distribution either makes sense or doesn’t make sense. Or if the flaps are too big on one side or the other. I’m trying to think of other examples, but, anyway, I think a good principle is that everyone should try to have a broad understanding of the product as a whole, and that’s very important. It makes a big difference.
So, other than the