Steven LEVITT: My guest today, Aicha Evans, has followed an improbable path, born in Senegal, trained as an engineer, she leads a billion-dollar company called Zoox. Zoox is making a radical bet on the future of driving. They’re developing a fleet of robotaxis that will pick up and drop off riders on demand. But one glimpse of these cars and you will know what I mean by radical. No steering wheel, no front or back seat, just a big open rectangle with two bench seats facing each other. They’re only testing these vehicles, but I want to find out how long it will be before they’re functional so I can buy one.
Welcome to People I (Mostly) Admire, with Steve Levitt.
LEVITT: Have you noticed that I almost never have CEOs as guests on this show? That’s because my experience with CEOs is that they won’t say publicly what they actually believe. They just spout the company line. I’m making an exception today because I’ve heard Aicha Evans is different and we’re about to find out whether or not that’s really true. And just a heads up. At a few points in this episode, you’ll be able to tell I’m wrestling with a head cold. I’m feeling better, but thanks in advance for bearing with me.
* * *
LEVITT: Do you have the sense that autonomous driving has turned out to be a tougher problem to solve than the engineers and data scientists expected? I feel like I’ve been hearing for a decade that we were right on the cusp of fully autonomous vehicles.
Aicha EVANS: I absolutely think so. As an industry we’ve underestimated the problem, but I’m one of those optimists — I actually think that’s a good thing because when you have a generational challenge and opportunity, if at the beginning of the journey you start talking about how complicated and difficult it’s going to be you might not try. You might not get going. But progress is definitely being made and I’m glad we all got onto the journey. I consider this to be basically something that’s inevitable. It’s a matter of time and it’s really going to change a lot of things in our lives and environment and cities and how we do things and how people move around.
LEVITT: Yeah, if this doesn’t happen then I think our entire vision of what the future’s like is very different than what we’re imagining. So first, maybe you could explain what the strategy is that Zoox is following and how you fit into the broader schema of autonomous vehicles.
EVANS: Yeah. So for Zoox, it’s really, we’re a transportation company. We’re setting out to reinvent personal transportation to make it safer, cleaner, and more enjoyable. We basically have a fleet of fully electric, fully autonomous vehicles that are available on-demand when you need them. And when you don’t need them, somebody else is using them. So, it’s extremely efficient from that standpoint. We handle the maintenance, the charging, the driving. You basically say, I want to go from point A to point B. We show up, sliding doors. Step in, sit down, buckle up, push start, personalize your space a little bit. And by the time we drop you off, we are picking up another passenger. So we are really all about riders and people who need to be transported from point A to point B. And that’s a very different model for most. It’s not like mobility-on-demand is a new concept, right? We have a couple of companies in the United States that are providing it today, but it’s with a human driver that is driving a passenger car.
LEVITT: It seems to me that there are three very different types of barriers that stand in the way of fully autonomous vehicles. One is the technical challenge, the science and engineering of solving a really hard problem. The second barrier is human psychology, getting humans to trust the technology and to change their habits. And the third is politics and regulation, because I think even if you have a great technology that people would use, powerful interests are going to try to stand in the way of actually letting us put those into practice. So, I was hoping you could just run through those one by one, starting with the technical side.
EVANS: Just getting a vehicle to drive is not that difficult, but getting it to drive amongst other human drivers and being able to follow its trajectory, get to the right place, follow not just the rules, but the norms, and then predict and react to what other drivers will do, which are, human drivers. And so, now with the advances in robotics, in A.I., and also in general in compute we do have the foundational element. It’s now about testing, having quantifiable ways to say, is it safe? And is it doing the right thing?
LEVITT: What can the technology do? And what can’t it do? Cause my guess it’s also the standard you’re holding it against — my guess is if you put the technology head to head with a human, it actually comes out pretty well even right now, but that’s probably not the standard you’re using to judge whether the technology’s ready for prime time.
EVANS: Exactly. So the technology does well in normal environments today. And it does actually better than humans. However, in terms of handling all of the long-tail type unpredictable situations — somebody jumps in front of you, somebody on the other side in traffic is not respecting the rules — being able to handle those scenarios over and over again in a reliable way, that’s what the technology is not fully capable of doing — yet. And that’s why we’re all testing and developing our safety models to be able to basically do that. So, it’s not about doing a route once, it’s about doing it thousands and millions of times in a very predictable way, no matter what’s thrown in front of you in terms of being able to predict it or react to it.
LEVITT: So, I presume these artificial intelligence systems that you’re using, they learn from experience. How do you train them? Are you letting them drive around all day with a human behind the wheel? Or can you actually build simulations that give the systems practice off-road as well?
EVANS: All the above. So three forms of testing. First, we have, in our case, Toyota Highlanders that are outfitted with the exact sensor and compute configurations, and they are driving around downtown San Francisco, as well as Las Vegas, shift after shift. And there is a safety driver and then when they come back to base we take the log and we review what happened. And if a driver took over we try and understand, was it just a precautionary measure? Or was the vehicle going to do something it’s not supposed to do or we don’t want it to do and then the driver took over? Then we have the purpose-built vehicle where there is no safety driver whatsoever. So we’re testing that on private roads, on very deterministic routes and starting to teach it. So, it’s not as advanced as the Toyota fleet, but it will get there. And then simulation, simulation, simulation, because not all miles are born equal and you want to be able to not just simulate, but add scenarios, add agents jumping in front of it, doing things you don’t expect. You apply effects, and then you do it again.
LEVITT: So, let’s talk about the human psychology, the fact that people distrust machines and that people like to feel like they’re in control and that old habits like driving are really hard to break. Do you think these human factors are going to slow down the path of autonomous vehicles more than the technological constraints?
EVANS: I don’t think so. At least, in our case. When you look at a passenger car, you look at the windshield, you look at the shape, the hood, the steering wheel, brakes, pedals, the control cluster for instrumentation, even the seating arrangement — the fact that the passengers are to the side of you and behind you so they don’t distract you. Taking that type of vehicle and saying it’s going to be autonomous is going to be very hard because anytime a human steps into one of those, they’ll be thinking about driving. In our case, we went with a purpose-built approach — we’ve been at it since 2014, and we’ve been super focused and super consistent. And we were like, “Okay, A.I. is going to do the driving, via compute and sensors — how do we really have the customer and user-experience design so that if you step into our vehicle and think about driving, we failed?” We want you to be transported. And if you do that well, and you go step by step, little by little, then you’re delivering value and human beings, in general, love technology if it’s delivering value. So, I think the customers will come.
LEVITT: I wonder whether — I never really thought about this before, but there are a lot of people who are terrified of flying. And one reason is probably because the plane’s way up in the sky and you can imagine it falling from the sky. Another obvious reason is that they aren’t in control. They’re relying on some pilot they don’t know to drive them. I wonder if your purpose-built vehicles will create a whole new kind of phobia — the phobia of riding around in something where you don’t have any control over it.
EVANS: Yes, it will. For sure. I mean, we could have waited a much longer time being closer to launch to basically show the vehicle and show what it’s capable of but we’ve started a conversation with select folks and we hosted quite a few families and had them sit in the vehicle, give us their feedback. We’ve also thought a lot about the customer experience. Now I’m a little biased. I will admit that, but I can tell you that when I ride in our vehicle for the first, let’s say, meh — five minutes or so, I’m like “Ma, wow, this is just amazing.” But you know what happens? After five minutes or so, I’m on my phone, I’m listening to music, I’m doing something different. I totally forget. Now we’re going to have to earn it at the end of the day. It has to be safe. And I’ve said publicly, unless I’m willing to put my own children in there, it’s not safe yet.
LEVITT: In 2018, I remember a pedestrian was killed by one of Uber’s self-driving cars. And she was one of maybe 40,000 people who died in car accidents that year. But the result was that Uber suspended test driving in Arizona and California. And also just to add, I looked up the data — there were 58 people killed in Uber vehicles driven by humans that year. And nobody really complained. It will be really difficult for autonomous vehicles to take-off if the perceived acceptable level of risk is zero, don’t you think?
EVANS: So, it’s something we discuss a lot in our industry and we don’t get to whine about it. It is what it is. You know, humans have tolerated norms of acceptance with certain situations. It took a long time to get to that point and from the beginning state, the expectation is going to be zero, which by the way, is not realistic. And so somewhere between 40,000 plus and zero, we’re going to have to build a relationship with early adopters and then create momentum and create movement. But I’m not one of those who’s going to sit here and tell you that it’s going to be zero. It would be disingenuous.
LEVITT: Why do you think that humans are so forgiving of the mistakes of other humans, but have so little tolerance when an algorithm makes a mistake? Because I think most economists are of the view, look, I care about the decision. I don’t care about the process. If the car drives the same as the human, but makes different mistakes, that’s just fine. It doesn’t matter. But I think that is not the way the typical person reacts. They will be livid about any mistake the algorithm makes that a human wouldn’t have made, even though the algorithm avoids a thousand mistakes a human would have made.
EVANS: I don’t think we’re just logical creatures. I think that first of all, we’ve been together for a very long time. There are norms around intuition. There is a difference between the brain and the soul. There are also emotions and that leads us to have essentially a reservoir around forgiveness, around expectations. We know ourselves that we’re not perfect. We take that for granted or at least those of us who try to be honest with ourselves. And when it comes to the machines, we’re like, “Wait a minute, you are supposed to be totally logical. You’re supposed to have an algorithm that doesn’t make any mistakes. And if there are some, you need to have error correction.” We don’t have an emotional connection to that and we don’t have a discount for that.
LEVITT: So, I will say I’ve never seen a piece of technology that I thought I needed or would help my life. I could not understand why I would ever want email — a phone? What would I do with a phone? Why would I want to have a laptop? But I would really want to have a driverless vehicle and I want my own driverless vehicle in my own garage ready to take me everywhere I want. So it’s funny. I must be totally out of step with society when it comes to technology.
EVANS: I don’t know that you’re out of step. The question I would ask you is: why do you want your own versus would you be okay if anytime you want to go somewhere, one is available to you?
LEVITT: Well I do kind of want to personalize it. I want a little cooler in there and I want it, you know, set to my stations. I want to make it messy and not have to clean up after myself.
EVANS: Well, those are two different things. We can definitely have a profile with your preferences and make sure that we cater to them. In terms of making it messy, I’m going to have to charge you because then the vehicle has to go back to base before it’s deployed again.
LEVITT: So, I’m curious — clearly in your model there’s the idea that car ownership will essentially disappear, but it’s not obvious to me why I wouldn’t want to be parking one of your vehicles in my own garage. But it doesn’t seem like that’s much of what you have in mind.
EVANS: No, especially for dense urban environments. We obviously are very mission-driven, but we understand we have to be realistic. I mean, I like to drive — sometimes, I live in California, I just go for a drive. The question is, what about transportation when you’re in cities? You’re going shopping, you’re going to work, or for entertainment, or whatever the reason is — if we can provide the service of transporting you from point A to point B and just providing that as a service, as we’re seeing today with mobility-on-demand, we think that there’s a market there. But it will be a long time before that is the only way of being transported. And as I like to tell people, when basically the horse and carriage were disrupted by the Model T, the horse and carriage didn’t fully go away.
LEVITT: So, your vehicles will be obvious and I wonder if they will suffer the same sort of fate that student-driver cars suffer now whenever mean-spirited people see those cars, they do what they can to make life miserable for the student drivers. Have you experienced that phenomenon with your vehicles on the road?
EVANS: Not to the extreme, but sometimes we receive an email or something and somebody says, “I was on the road and I did XYZ. And your vehicle did ABC.” Now what I invite everybody to remember is that in our case, we have sensors on all four corners of the vehicle and we have 270-degrees at every corner. So we have, basically, varying 360-degree field of view. And that’s camera, LIDAR, and radar. So remember we are recording everything all the time —.
LEVITT: That’s good.
EVANS: That’s all I have to say to those mean-spirited people who want to test us.
LEVITT: I love that answer. Okay, so the third obstacle to self-driving cars, even with great technology and customer acceptance, is political. There’s something like 4 million people who drive for a living. That’s 2.5 percent of the U.S. labor force. My guess is they will fight like crazy against driverless vehicles. And also if I’m a regulator who will be skewered if a single self-driving vehicle kills one person, I’m definitely in no hurry to start changing the rules about making it easier for those cars on the road. So, tell me about that. What’s your view on what will happen with this political fight?
EVANS: So, there are layers and layers and layers here. So first of all, this is going to happen either in the United States or elsewhere. Second thing is we talk a lot about the human drivers that maybe will not have jobs because they can’t drive anymore. And I’m talking decades and decades. None of these transitions happen overnight. It takes a lot of time. So, let’s talk about all of the jobs we are going to be creating — every time a new mode of transportation occurs and takes hold, there’s an infrastructure around it. The vehicles have to be serviced. They have to be charged. And so we will bring new types of jobs that will also be available to maybe people that were driving before. And because the transition will be slow, there will be plenty of time to basically make the adjustment. Then there’s the whole infrastructure issue. Our cities today are plagued with congestion. We know that a lot of the pollution and a lot of the harm to the environment is done by essentially transportation with the internal combustion engine. And we also know that most of our cities today, especially the big ones, are already built. You can’t really redesign them or reconfigure them very easily. So, we think that if we are careful and we take it step by step, there are enough forces that are coming together where, in aggregate, society benefits and when society benefits, that’s the higher order bit for politician and from a regulatory standpoint.
LEVITT: Yeah, that makes sense. One of the things that is incredibly alluring to me is the idea that once we fully transition — so human drivers are completely out of the picture — then the ability of an armada of autonomous vehicles, communicating with one another totally changes the dynamics of the road. So on a freeway, I don’t know what the numbers are, but you’ve probably seen studies that would say that we could have three times, four times, 10 times as many cars on the same roads if we didn’t have to have such huge buffers in such differential speeds of the different vehicles.
EVANS: We’re talking way in the future. But you are absolutely correct. When we get to the point where certain areas, certain zones, or certain chunks of a highway are fully autonomous, you get the joy, from an algorithmic standpoint, of controlling throughput. And when you can control throughput in a very deterministic way, then you can stuff the pipeline way more than when you have a bunch of undeterministic agents driving at different speeds and doing different things. And you don’t really know where they are going and what they will or won’t do. This is probably decades away, though, we already see signs of it. For example, in Beijing they have the system of the rings. Not all cars are allowed to go to all rings of the city. So, I would not be surprised if we start seeing, in actually a decade or so, specific geo-fenced areas where from a city or state standpoint, they decide that area will be fully autonomous. And then we’ll be seeing the benefit of throughput and pipeline optimization. But at scale, it’s gonna take decades.
LEVITT: So, take San Francisco. If you had to make a guess, what year will it be that 90 percent of the cars on the road are autonomous — fully autonomous?
EVANS: I’m not very good at guessing or gambling, so I’m going to be careful here. I’m going to stick to decades.
LEVITT: You really think it’ll be 40 years before that’s true?
EVANS: First of all, decades, that can go all the way to 90.
EVANS: So, how about I’ll make you a deal — between 50 and 90? Yeah?
LEVITT: Well, unfortunately, I’m not going to be around to experience it, so I can’t even bet you on that. It just seems like it should happen — I mean, look at what happened with cell phones. We went from carrying around things that were like, huge rocks and couldn’t do anything to like if I didn’t have my cell phone, I would want to kill myself, in about a decade. It’s obviously a different problem, but I’m surprised you think it’ll be such a slow transition.
EVANS: When I talk about San Francisco, I’m talking all of San Francisco, meaning basically the entire San Francisco little peninsula. When will we see areas of San Francisco — where that happens? Much sooner. There I would guess a decade or two max.
You’re listening to People I (Mostly) Admire with Steve Levitt, and his conversation with Aicha Evans, head of the autonomous vehicle company, Zoox. After this break they’ll return to talk about the economics of the autonomous vehicle industry.
* * *
LEVITT: Hello, Morgan.
Morgan LEVEY: Hey Steve, how’s it going?
LEVITT: It is going great. And if we’re hearing your voice, it must be Q and A time. So why don’t you fire away?
LEVEY: So, at the end of our episode, from a few weeks ago with Maya Shankar, you shared a hypothesis. Do you remember what that was?
LEVITT: I don’t have a very good memory, but I do remember that I made a conjecture that there was an inverse relationship between the need for control in one’s life and how happy somebody is.
LEVEY: Inverse relationship. You’re such an economist. So, you mean that less control in someone’s life means they might be more happy or if they’re more controlling, they might be less happy.
LEVITT: People who need to control their life tend not to be very happy, is my experience.
LEVITT: People who accept whatever comes, tend to be pretty happy.
LEVEY: You also said that you had changed in your own life, that you were able to let go of control and you were happier now than you used to be. So a few listeners had a followup question, specifically a woman named Linda wrote in to say that she would love to hear you elaborate on how you were able to let go of the need to feel in control all the time. Do you have a good answer for her?
LEVITT: So, Linda, I do have an answer to that question, although I think it’s going to turn out to be a little bit heavier than you hoped or thought. But, here’s my answer. I can think of two distinct events in my life that have profoundly affected the way I think about control. The first of these was a month-long trip I took to India and I talked about that at length in our episode with Sam Harris. So I’ll just leave that to listeners — if they want to hear that story, they can go back and listen to that old episode. But here’s where it’s going to get heavy — the other experience in my life that deeply affected the way I think about control was by far the most tragic thing that’s ever happened to me. I had a son named Andrew. He died suddenly, nine days after his first birthday, from meningitis completely out of the blue. And I had always feared something like that — in some sense losing my child was probably the deepest fear that I had. And I wish that I could say, “Oh, the reason I was able to let go of control is that my worst fear came true and it turned out not to be that bad.” But actually it was the opposite. My worst fear came true and losing a child was so much worse than I ever imagined it would be. And really the only escape from that for me was surrender. It was surrender to the universe and it was just, the pain and the loss was so great that I just kinda gave up. And I don’t even know if that will make sense to people listening, but to move on in life, I just gave into it. I just gave into the idea that I had no control, that I was nothing, that the world was going to do what it was going to do to me. And I had no choice, but to accept that. Look, and there was virtually nothing good that came out of his dying but I have since then been more or less free of the need for control. And I wish it could have happened in any other way than the way it happened.
LEVEY: I know that you really like the serenity prayer. Is that why?
LEVITT: Absolutely. So I don’t even know the serenity prayer except for the first two lines. “God, grant me the serenity to accept the things I cannot change, the courage to change things I can and the wisdom to know the difference.” And I think those are some of the most profound words I’ve ever read.
LEVEY: Yeah, me too. Thanks for sharing, Steve. I hope some of our listeners have found some comfort in what you said. If you have any questions for us, please write in our email address is email@example.com. We’d love to hear your thoughts, ideas, and questions about anything. Let’s get back to the interview now.
* * *
LEVITT: Where do the existing auto manufacturers stand on the issue of autonomous vehicles? I suspect that they’re pretty happy with the status quo, but I also suspect that they want to feel like they’re going to be part of the future, although they’re probably not very well positioned to do it because it’s really hard to defend the status quo while bringing in the future. Are they fighting it? Are they part of it?
EVANS: I think it’s a little bit of both. Very traditional when it comes to a core business that’s doing well but that’s also being disrupted. First, in terms of the status quo, they are trying to make the current vehicles as sophisticated as they possibly can. And also the E.V. transition is definitely picking up steam. When it comes to autonomous driving in terms of a fleet of robo-taxis, that’s just for transportation and pick-up and drop-off, you see a little bit less momentum there, but that’s to be expected. This is why most companies are very successful and then they basically sit at the top of that success. And then they forget to look at the disruption that’s coming, either sideways and from the bottom. And then a lot of them, over a cycle, disappear because they miss the transition. Again, this is inevitable. So, those who embrace the new world will be fine. And those who don’t will have a very difficult time. This has been true since the beginning of times. Doesn’t matter how well you’re doing, at some point there will be disruption. You embrace it and you get to be around for the next round. You don’t embrace it and you go away.
LEVITT: Yeah. This is an aside, but do you think it makes sense that Tesla has a market-cap of over $500 billion more than the next six auto manufacturers combined? I don’t know anything about it, but it seems really bizarre to me.
EVANS: Ah. That’s a tough question. Look, I don’t know that I’m qualified to say whether it makes sense or not from just an absolute value standpoint. There are a lot of things in this market that are quite mind-boggling, shall we say? But I want to give them credit. At the end of the day, they are disrupting the industry. At the end of the day, they have found a way to do it while developing technology and participating in the innovation while already delivering a product to customers. And sometimes — I’m sure maybe my P.R. team will be like, “What did you say in a podcast?” But the reality — I think fair is fair.
LEVITT: So, Tesla and Zoox are taking a really different approach to autonomous driving. So, they’re incrementally trying to implement within the given system, ways of building in levels of help from algorithms. And you have taken the incredibly bold strategy of skipping over everything intermediate and jumping right to the punchline. So, let me just say what seems so radical to me. The first one is that you’re building vehicles that don’t have any space for anyone to ever drive. So you’re not hedging against the possibility that you’re going to need somebody sitting in the front seat. Which makes total sense to me, but it’s risky, right? It’s a big swing. Secondly, you’re not content like, say, Uber or Lyft, to be providing the taxi service. You’re not content like Tesla or Ford or General Motors to just be making the vehicle. There must be people who are just in the software part of it as well. It just seems like a really big — thinking big, which I’m all in favor of, but I don’t think you really captured how visionary what you’re trying to do is. You’re probably trying to be modest or something, but it seems to me like succeed or fail. It really takes a lot of chutzpah to be trying to do what you’re trying to do.
EVANS: I haven’t heard the word chutzpah for a while, so thank you for that. That’s making me smile. Look, we actually look at it very differently. We look at it as we are the most focused of everyone. We are taking the subset of the different technologies, vehicles, electric vehicles, robotics, A.I., the service, software simulation, we’re taking the intersection of those that is required to solve the problem and solve it correctly. We don’t think that taking a bunch of cars and retrofitting them by just slapping sensors on them is going to do it. And so we think that you architect the vehicle with that intent from the beginning and we have not deviated. And we’re not doing anything else. We’re not doing trucking. We’re not doing logistics, we’re just doing this people mover. And yes, it has no manual controls because, as we started talking about earlier in the podcast, we don’t want you to think — you, the customer — to think about driving. We’re going to transport you. And so we think that we have an advantage when it comes to iteration and speed, because all of the disciplines are in one room or one company, you don’t have five different companies with different pedigrees, with different interests, with different intent, trying to slap things together. And then in terms of people, we started with two co-founders and that was it in 2014. We’re up to about 1,200 people and hiring, by the way, in all of these disciplines. Everybody looks at it as we’re trying to do a lot. We look at it as we’re super-focused and we’re using all of the pieces, but only the necessary subset that is needed to integrate, collaborate, get to market. And at the end of that is a fleet of robo taxis that are fully electric and fully autonomous. And all you do is ride.
LEVITT: When do you hope to have the first paying customer climb into one of your vehicles?
EVANS: Aha. That’s the most important question that I don’t give an answer to. No, I’m teasing. Look, the answer is not next year for sure, but certainly much sooner than people realize. We have line of sight. There’s still some elasticity because we have some important work to do, but it’s not next year, but not too far in the future either.
LEVITT: I would love to talk about the economics of your business. In the regular taxi business, labor’s a high component of costs relative to the vehicle itself. And in your business model, you save enormously on that dimension. But, at least for now, I suspect the cost to produce one of your vehicles is sky high. Is that true?
EVANS: Let’s just say that it is not inexpensive to produce one of our vehicles. Yes. But remember, we’re monetizing it ride by ride. And this is an electric vehicle. This is not like a passenger vehicle that is essentially built with the understanding that it will be idle 96 percent of the time and only driving 4 percent of the time. These machines are built to be driving all the time with a very long useful life. I don’t necessarily just look at the absolute value of the cost that is being added. I look at how much is that going to add per ride? For example, today when it comes to pick up and drop off, the statistics or the information says that about 40 percent of the time mobility-on-demand drivers are basically driving without a passenger. We can optimize that because we don’t go to get a coffee. We don’t go to the restroom. We don’t decide to stop for a phone call. We’re just picking up, dropping off, picking up, dropping off. We know where the next passenger is and we can optimize the placement of the fleet. And so we look at cost in a very different way. Now, in absolute terms, I want this vehicle to be as inexpensive as possible because that’s just pure margin, but we’re a lot less sensitive to it than most would think from a business-model standpoint, because it’s really how much costs are you adding to a ride?
LEVITT: But that high cost is the reason why you don’t want to sell me one of these, even though I really want to buy one and put it in my garage. That’s the reason why you’re not thinking that direction.
EVANS: You are too smart. First of all, it’s not practical, but yes, it wouldn’t make sense.
LEVITT: So, I wrote an academic paper using Uber’s data and two key insights emerged from that study. And the first one was that the consumer surplus associated with ride sharing was huge. In other words, the benefits that consumers got from the existence of Uber being available was far greater than the amount of revenue the company was getting, or that the drivers were capturing. Basically, it was just saying, this is an amazing innovation for society. And your business model is exactly along the same lines and would also yield these huge social returns. But the second insight that emerged from that paper was that by any reasonable standard, Uber was charging way too little for its rides. So for the economic nerds out there, Uber was pricing on the inelastic part of their demand curve. And you learn in Econ 101 that a profit-maximizing firm should never price on the inelastic part of their demand curve. I think partly Uber was making a mistake and they didn’t realize how poorly they were pricing, but also I think they were caught in a very competitive situation. Tough competition led them to want to charge low prices, to try to capture market share. And I wondered, do you ever worry about the possibility that you’ll overcome all of these other obstacles and then cutthroat pricing will leave you with no margin in the end?
EVANS: So, I have to be careful here because I also don’t want to come off arrogant. That would be bad. I worry about it in terms of I monitor for that tendency. But I’m not too worried that we’re going to do that because first of all mobility-on-demand, a contract has been established, that’s one, the second of all look, this takes a lot of technology. This takes a lot of talent. You can’t just wake up as a willy nilly company and say, I’m coming into autonomous driving tomorrow morning. It takes time. We’ve learned a lot — several generations of iteration. So, we think that the competitive landscape is going to be a little bit simpler for us because of that. We’re not saying we’re going to be the only ones. This is a multi-trillion dollar opportunity according to Morgan Stanley. And we know that those don’t get controlled by one company. That’s just not the way it works, but we think that the time to be low-barrier of entry and very commoditized is many, many, many, many, many decades away. Now, will there be price war, especially in certain markets? Probably, but you know what, that’s okay, the customer benefits. At the end of the day, we think that there’s a lot of elasticity in the business model and we are very resilient to a variation in pricing.
LEVITT: But overall your plan is not to follow the Uber path. So far they’ve run up over $20 billion in losses. So you’re not hoping to follow that path I take it.
EVANS: No I am not one of those who puts a profit ahead of everything. You build great technology, a great company, a great culture. You provide a product that customers want and then out of that the outcome will be a great business and great business equal, revenue, margin, net income, free cash flow.
LEVITT: You had an incredibly successful career at Intel rising through the ranks and then becoming the chief strategy officer. Do you ever think you were crazy to leave to come be the C.E.O. of Zoox?
EVANS: Oh, that’s a funny question. I assume you want an honest answer, yeah?
LEVITT: Oh, of course. I’m always looking for honesty.
EVANS: Look, yes and no. Something happened to me the last few years where I was very happy at Intel. It’s a company that I love, but at the same time, I had been there for a while. I was getting a lot of phone calls to do something different and I said, “Ah, I’m not ever leaving Intel to go to another big company. I’m happy here.” At the same time you start getting to this point in your career where obviously you want to be compensated fairly. You want challenging assignments, but you start thinking about meaning, impact, legacy. And when the call came about Zoox, first immediate reaction was like, “Oh, that sounds complicated.” It’s very worthy. And remember, by trade I’m a wireless engineer. And I think that wireless is also a transportation business. We just happen to transport bits that lead to access and knowledge. And obviously we’ve seen what happened with wireless over the last, what? 30 years? It’s been amazing what it’s made possible for all of us. So I talked to board members at Zoox and I talked to our co-founder Jesse Levinson. And obviously I got to see the company from the inside and within about 30 minutes of the onsite visit, I was like, yeah, right on. Legacy, meaning, changing the world. I maybe won’t be here for when Zoox is deploying in Africa, but it’s still worth it. But it took about a couple of weeks of a very isolated vacation with my family to pull the trigger and say, “Yeah, let’s go do this. You can afford it. And best case it works out and you do something super meaningful and super transformative and super impactful to society. And worst case you learn a lot. So why not? Let’s go.”
LEVITT: Let me ask you about failure. So, as a manager, I would assume one of the most challenging issues is that ex-ante, you want people to take risks and that will inevitably lead to failure, but ex-post, there’s a real tendency to blame people for failure and often failures result because people did the wrong things or made choices that weren’t great. How do you manage failure at your company?
EVANS: So first, you have to state that it’s okay to fail. And then when there’s the first or second failure, you have to control yourself because you can’t tell people it’s okay to fail, and then you’re like the meany and out of control. It doesn’t work that way. So you declare it, you own it. But then you have a few rules. If you talk to somebody at Zoox, I’m pretty sure they would say, “Oh yeah, she says, basically make new mistakes.” If you just live by that, then it becomes really calculated risk. Second of all, I give up to three times — by the third time, I’m like, “Okay, we have a problem here.” And then last but not least, when you fail, the most important thing is really the learning. What happened? Why did it happen? And then how can we make sure that it doesn’t happen again?
LEVITT: Are you able to cultivate a culture at Zoox where people will tell you you’re wrong? Or do people just, yes you all the time?
EVANS: The majority of people yes me all the time, until the anonymous survey comes in. That’s the normal methodology. But it is my responsibility to have a set of people — it’s fairly small, who will tell me that I am not queen of the anthill, who will tell me that I’m wrong, who will tell me that I’m inconsistent. There’s one person on the team. She starts with, “Miss Aicha, we need to talk.” And so there are about, I want to say 10 people or so, plus my co-founder, we talk all the time. We’ve built in two and a half years an incredible amount of professional intimacy and support for each other. And then there is my husband and my two kids. They will definitely tell me.
LEVITT: It’s shockingly rare to see African-American women running big companies. I think in the 66-year history of the Fortune 500, there have been three African-American women C.E.O.s. Do you feel any responsibility to fight for change in the system or do you see your job as just to be the best C.E.O. you can be and leave social change to others?
EVANS: It’s evolved. Yeah, I would own the fact that initially, I was all just about being the best that I can be. And I want to say over the last five years or so, through some people talking to me, I’m going to admit, and also through just observing some things, I feel a little bit more of a responsibility. I don’t want to fight for change though — I want to invite and I want to incentivize, and I want to create an environment where it becomes abundance for everybody. I really don’t want to live the rest of my life being angry because somebody did this or somebody did that. It’s just not a very, at least for my personality, that doesn’t work. There are a few things that have happened to me where I’m like, “Whoa,” it is important that people see that it’s possible, that people see that I’m normal, that I provide or help with networking with opportunities and so on and so forth. But I think it’s also important to — look, on the one hand it’s been difficult for me because I’m an African-American woman in high-tech, but the reality is on the other hand, I’ve also had opportunities because I’m an African-American woman. I trusted early on, I will never forget, a young engineer in — oh, well he was older than me back then — fairly young engineer in Austin, Texas. We were working together and I was the youngest one out of the group and you’re always like, “Oh my gosh, I need to do a good job” and everything. And he was like, “Aicha, you can always speak up. First of all, you’re an R.C.G., recent college graduate. We don’t expect much. And so if it’s dumb and stupid, it’s like, you don’t know any better, let’s teach you. And if it’s interesting, we’re like, whoa. And it’s very likely that it’s going to be a perspective that is very different from any other perspective that is offered in the room because you come from a different environment and a different upbringing. So please trust that.”
LEVITT: So, you were born in Senegal. I’m curious whether you were extremely driven and single-minded in your focus right from a very early age? Were you different from the other kids?
EVANS: Well, I mean, okay. My father was in telecommunication engineering. So that’s what I knew. I had to hack a rotary phone to have access to my friends. And so a few things that looking back — first of all, in my family, it wasn’t really an option that you’re not going to get educated. The alternative wasn’t presented. It’s just the expectations that were set. And what became normal was you go to school and you get a degree and you work hard and you become a functional adult.
LEVITT: Hacking a rotary phone? What did you mean by that?
EVANS: When I lived in Paris, I would go back to Senegal and I would want to call my friends. And so we had these rotary phones and I was calling my friends in either direction and very, very expensive bills. And we were doing well financially, but my dad was like, this is ridiculous. And so he put a lock, there’s a little lock that you could buy, it literally had a key and you would turn it and then you couldn’t move the little wheel to dial numbers. What he forgot is that you have the handset and when you take the handset off the base, there are two little clicks. So I opened up the phone and when you’re dialing, what’s really happening is that the movement of that dial, depending on the delta T distance, provokes, basically a number of clicks on those little clicks. So, I figured out how to write down a phone number in terms of number of clicks with the right interval in between.
LEVITT: Would you dial wrong numbers pretty frequently?
EVANS: Yeah. It was a catastrophe at the beginning. Because getting the click is easy, but the space and time between the click, it takes a little bit of time to master that rhythm. But eventually I got pretty good at it. And I would dial, with my finger and I would tap — one would be one click, two is two clicks, three is three clicks. And I figured out the rhythm perfectly. And my dad was like, “I can’t understand this, we’re still getting crazy phone bills. And it’s the same phone numbers.” And I’m like, “Aha. Aha.” He still jokes about it.
LEVITT: It was probably one of his proudest moments I would suspect.
EVANS: Back then he was very unhappy with me, but I’ve started sharing this story publicly the last couple of years. And he said, “Now I can tell you, that’s how I knew you were going to be okay.”
LEVITT: That is awesome. I love that. So what advice do you give to young people who are listening to our conversation? You’ve obviously been incredibly successful, but you’ve also kept a spark and a joy around you, which is really unusual I have to say. So I’m just curious, I mean, I look at you and I think, wow. If my daughters could be like you, both with the success, but even more so with the spark — so let me forget about success. I don’t care about success. The spark. How do you keep the spark?
EVANS: Happiness is a choice. That’s one. Second of all, I just feel that in the grand scheme of things, life is really good. And somebody who is inquisitive, somebody who takes a lot of risk in life, I enjoy that. I enjoy trying new things, seeing new things and really exploring what I’m interested in. And it’s almost like I have no other choice, but to be happy about it because otherwise, the alternative is I’m miserable. And if you’re miserable, then you have fear. I mean, when it comes to Marie Curie, when she says, “Nothing is to be feared, it’s only to be understood and now is the time to understand more so we can fear less.” I truly believe in that. And it’s just a practical thing. Look, if you give me business assignments, that’s very steady state and keep the business running and maintaining, and you really need the revenue out of it. You shouldn’t give it to me because that’s not my personality. That’s not how I get into my happy place. I like things where you’re not sure is it possible or not. It takes a lot of risk to figure it out. And I accept that with that will come failure. If it’s easy all the time, that means we’re not pushing the envelope. We’re not innovating. We’re not taking risks.” So if you want to learn to live with failure, because it feeds another happiness in you, which is the happiness of discovery and creativity, you have to work on staying happy.
LEVITT: Choose happy. Simple words of advice from Aicha Evans. Is it really that easy? I’m sure not. But, on the other hand, it’s a life strategy that is so obvious that it never really occurred to me. Can I really choose happy? That would be amazing. Well, it definitely deserves a try, don’t you think? And if you give ‘choose happy’ a shot, let me know how it works out for you, good or bad. We always love to hear from you about that or anything else. The email address is firstname.lastname@example.org. Thanks for listening.
* * *
People I (Mostly) Admire is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and Freakonomics, M.D. This show is produced by Freakonomics Radio and Stitcher. Morgan Levey is our producer and Jasmin Klinger is the engineer. Our staff also includes Alison Craiglow, Greg Rippin, Joel Meyer, Tricia Bobeda, Emma Tyrrell, Lyric Bowditch, and Jacob Clemente. All of the music you heard on this show was composed by Luis Guerra. To listen ad-free, subscribe to Stitcher Premium. Thanks for listening.
LEVITT: Does Amazon leave you alone or are they up in your face all the time?
EVANS: I don’t know what that means.