Uber 创始人 Travis Kalanick 谈 AI、风险与打造未来 | EP #164

Will robots and AI replace all of our jobs?

If somebody wants to go and make a thing that doesn't make people's lives better, they should count on failing.

In your 7-year tenure: 0 to 70 countries, 5 billion rider trips, $7.5 billion in revenues, and a $70 billion valuation.

I didn't think of it exactly like this at the time, but basically, I was digitizing the network for the physical world. When you're trying to do something disruptive, there's a massive immune reaction that occurs. One of the best ways to deal with resistance to change is to build a massive amount of trust with those you're asking to change, or that you're asking to bring along with that change, because then you can turn your adversaries into advocates.

Nobody's in control of the world. So, you have to see where the world is going and build for that world.

Now, that's a moonshot, ladies and gentlemen. This is a group of entrepreneurs running $10 million- to $10 billion companies. Their focus is on getting clarity on their massive transformative purpose and then using that for a moonshot to do something extraordinary—to go from success to significance in the world.

When we've talked about understanding the process of reinventing an industry, transforming it, and making it efficient, I think there are only a handful of individuals who make that possible or have done that successfully. I want to read the numbers on Uber, just as an example, because I want to dive into the founding of Uber and then, obviously, what you learned in building CloudKitchens as well.

In your 7-year tenure, 0 to 70 countries—that's pretty impressive—5 billion rider trips. I want to dive into how you got started. Can you take me back, first of all, to that founding aha moment, and then we'll talk about purpose as a critical part of building a moonshot company?

Yeah. I don't know where that number came from. It came from ChatGPT or some site like that. When I left in mid-2017, we were probably doing—I don't know—13 million a week.

Okay. So, that would be like 7 weeks for every billion. So, maybe that's $7.5 billion in revenues. Not bad, thereabouts.

Yeah. Well, the gross at that time was probably about $50 billion. That would be GMV, and then how you calculate revenue from there could be tricky, but yeah, a $70 billion valuation—not bad.

The aha moment is kind of interesting. You hear a lot of founding fables. The public sort of generates this whole made-up thing. We don't have one of those, but it sounds like one.

We were in Paris. We were at—if you guys remember way back in the day, in 2008 and 2009—Loïc Le Meur's LeWeb. I was sort of in between gigs. I had sold my recent company to Akamai, and I didn't know what was next.

I was there with a guy named Garrett Camp, and we were walking back. Before Uber in Paris, you'd have to walk back. You would go out to dinner, and you'd walk home. That's just what it was. We were on a 3-mile walk back to our hotel after dinner, and he said, “Man, I wish we could just push a button and get a ride.”

That's a prophetic wish.

I'm like, “Pretty good.” He was like, “Yeah, let's just go back to San Francisco, make the app, buy 20 S-Classes, get 40 drivers in a parking garage, and we're off to the races.”

I'm like, “Look, there's enough cars out there already. We don't need to buy the cars. We don't need to get a parking garage, and we probably don't even need to hire the drivers.” That was kind of the balance at the beginning: He was the classy one, and I was the efficiency one, if that makes sense.

Purpose—we talked about this in advance. How critical is purpose in driving a company at this scale?

It's super important. You could probably have a whole conference on purpose. You could take it many different ways.

The first is, I think every individual—sometimes I'll talk to executives who are making the move to the next thing, ideally the one I'm trying to recruit them for. My advice, even when I'm not recruiting them, is that you've got to be really self-aware of who you are. If you get really, really self-aware of who you are, then when the next thing comes, you will know when your professional soulmate presents itself. You will know.

Fascinating.

If you're honest with yourself, you've got to be. It's like, I shouldn't be running Pinterest. It just wouldn't work out, but that's because I know myself.

If you understand your nature, you need to find a thing that matches that nature. So, when we talk about your massive transformative purpose—what wakes you up in the morning and what keeps you going—as the foundation for building a company, how does that translate to the team you build?

Before we even get to that, there are a couple more layers to it. I like to say that my sort of space is digitizing the physical world. Let's say that's the sport that I play, that I'm meant to play. If you play basketball and that's your thing, and you're really great at it, don't go play tennis.

So, you've got to know the sport through that self-awareness. You know your sport: digitizing the physical world. I'd also say innovation at speed and at scale is the stuff that gets me fired up.

The final layer is: Who are the people that I'm serving? Whatever you're going to do, you're serving somebody. You should be fundamentally passionate about them and what the thing you're going to do is going to do for them.

No, and I think that's when, in the nomenclature here, we talk about your MTP: Who are you serving? Who do you want to be a hero to?

So, you get clarity on that. Then you start building an organization. I guess the challenge is, if you're bringing in people who don't share that, you're going to very quickly start going in the wrong direction.

This is when you get an internal culture clash: There are people who don't see things the same way you do, at least in terms of values. Most of us here have personal relationships of one kind or another—friends, loved ones, partners in life, the whole thing. Opposites can attract, but the core values need to be shared.

How do you use that as a filter for who you're bringing in? Are you dogmatic about it, or are you allowing those who naturally gravitate that way to come in?

Those 2 are, I think, very similar. Sometimes somebody who shouldn't be there is like a tennis player who goes to a basketball court, and you're like, “Whoa, whoa, we don't do rackets here. We have this much larger ball that we put in this hoop. It's different.”

It doesn't have to be a judgment. It's just an assessment, a factual sort of thing. You have to understand what the values are and whether the values are the same. Again, it's those things that I just went through.

This gets codified into what you might call cultural values or things like that. If your cultural values are really, really good, then it becomes really clear who's not supposed to be there.

You don't want to get to that place where somebody's not supposed to be there. When somebody comes in, you want to make it work. It doesn't always work, but you want to try to make it work. Ideally, it becomes really clear in the recruiting process, and if a mistake happens on the way in, it becomes clear super early: “Well, I play tennis, not basketball.”

You said a few minutes ago that you are focused on optimization and getting rid of inefficiencies. Is that your superpower for finding opportunities?

It goes to that digitizing the physical world. We know what the digital world is: It's called a computer. The CPU manipulates the bits, storage stores the bits, and the network moves bits from point A to point B. These are the 3 core computing resources in a computer.

But if digitizing the physical world becomes your sport, then you've decided that treating atoms like bits is what you do. So, you go, “Okay, well, the CPU manipulates the bits. What manipulates atoms?” That's manufacturing. “Storage stores bits. What stores atoms?” That's real estate. “The network moves bits from point A to point B. What moves atoms?” You're like, “That's logistics or transport.”

My last company, Uber, I didn't think of it exactly like this at the time, but basically, I was digitizing the network for the physical world. Digitized transport. One of the 3 core computing resources in an atoms-based computer.

You're like, “Well, that's well on its way.” It's not all the way done. We need autonomous cars and all of this to really be out there, but there's light at the end of the tunnel. What about digitized manufacturing and digitized real estate? That's how I see what I do.

So what I do is I make atoms-based computers. If you are treating atoms like bits, it will naturally be about efficiency. It's very different than, say, “I make a social media app.” A social media app is going to win when it takes your attention and your time, and that's how they make money. Atoms-based computers typically are giving you your time back.

Yeah. Now, we talked this morning about the idea of looking into your business and finding everything that's not been digitized yet and dematerialized yet. Somebody is going to eventually do it if it's doable.

Yeah.

What advice do you have for folks about doing that? The other part of it is companies that are born as digital-native companies, like you built Uber, and those that are trying to retrofit themselves.

Well, I mean, these are very different things. So, if you are digitizing a business that exists, let's assume that business is profitable and has some sort of competitive moat, et cetera, in some ways, the starting point is very straightforward. It's like you just take all the line items where there are humans, and you try to find ways to get more leverage on those humans.

So, you start with customer support, account management, any workflow where people are doing things, filling in data—workflow-type stuff. That's the easy starting point. It's not as innovative, per se; it doesn't get to the core product, but it's sort of how you start surrounding the core if you have something that's not digitized.

But the radical thinking is to really digitize it from the inside out. That's not always the domain of an existing profitable business with a competitive moat. It usually doesn't have the DNA to do it. Sometimes it does, but most of the time it does not.

I mean, we didn't see very many other preexisting transportation companies copy what you did, even though you gave a very clear model. Timing on this: I know Bill Gross was here with us today, but he spoke yesterday morning about one of the talks he's given, looking at the most successful companies and the companies that failed—whether it was their CEO, their capital, and so forth. His conclusion was that it's timing.

Being too early is identical to being wrong.

It's worse.

Yeah, because you waste a lot of money and time. I mean, it's the worst.

The company I did before Uber was networking software. I sold it to Akamai, but I was probably wrong and too early, which is really the worst. For the first 4 years, no salary. I didn't have anything, right? I like to say, I actually at some point had socks that said, “Blood, Sweat, and Ramen.” It will grind you to dust if you're early. And if you're early and wrong, that's intense.

You hit a key point when the theme of this session is technological convergence, but your convergence on the smartphone, Google Maps, GPS, and the 2008 recession. How important was that recession for getting the gig-economy people willing to drive?

I don't think it had much to do with it.

Was it a concept Uber tried before you? Did some people try?

I mean, the main thing that existed before us was a couple of taxi apps.

Mm-hmm.

But the problem was it was just a broken model. A lot of times, when people see an opportunity, they go after it the wrong way. Taking a slice of the taxi market: A, it's sort of very small. And B, you're on somebody else's platform, and they can squeeze you. You're basically getting yield optimization from the 20% that doesn't matter.

So, if a taxi guy agrees to go pick somebody up but then gets a call that he doesn't have to pay any commission on, or somebody waving their hand and hailing him, you're done. The reliability and the quality of the product you're going to be able to offer sucks, which you could say is similar to where online delivery of food is right now. Uber Eats or DoorDash on a restaurant is like a taxi app. You're getting yield optimization on a thing that's built for something else.

I remember my first Uber ride in San Francisco, when you were just running the black-car service. And someone said, “Oh, I was going from a party—let me show you this new app.”

Yeah.

And what's interesting was you didn't try to boil the ocean in the beginning. You began really with the black-car service in San Francisco. Can you talk about staging a moonshot start—staging something that would scale? Just that mindset of, “Let's get it working here,” right, before we grow it.

Well, look, we were in San Francisco for a year before we went to our second city. Our second city was New York. We didn't even do the low-cost product, what we call UberX. We didn't do that until—that must have been early 2013. Maybe mid-2012 was sort of halfway, and then early 2013 was all the way. So, yeah, getting it right matters.

In my current company, there are a couple of situations where we went too big too quickly because you can learn the wrong lessons from the last thing. I'm like, “Oh, this scale is easy. It works.”

So, I did this thing—the acronym was PMD. Basically, I didn't want to have happen at my current company what happened at Uber, which was that we invented this thing and then had copycats everywhere, in different countries and continents, that we had to then go fight because they copied our thing before we got there.

Yeah.

Even though we went to so many countries so quickly, we got there after the clones got there.

Right.

So, we didn't want that to happen again. I came up with this acronym. It was called PMD: parallel multicontinental deployment. It sounds like a nuclear war.

And we did it, but we went to some places we shouldn't have gone. It just so happens you can't make money on kitchens in India. It's just not possible. We also found places that were amazing. We went to Kuwait, a lot of places you wouldn't go, that ended up being—we went to Saudi, we went to lots of places that we wouldn't have otherwise gone without that mentality.

But, yeah, don't go to Indonesia, don't go to India, and don't go to Colombia. Those are the main ones. I remember when France made Uber illegal. My motto was, “Don't start a tech company in a country that made Uber illegal.”

How important is getting the system and the operation working really buttoned down before you start moving to additional new territories?

Yeah, I mean, there's a balance. There are a couple of initiatives that I have right now that are—you know, they talk about the moonshot inside the moonshot. There are things like that that make the overall thing win. We've got a couple right now that are very much in that category.

That balance between when to cook it and when to scale it is a tricky one. There's judgment and instinct, but this is kind of tautological and kind of obvious, so I'm not sure it's going to be helpful: You don't want the overall effort of winning to be dragged down because you went too big.

Okay, obviously.

Right now, I have this one project where it's still operationally intensive because we haven't gotten the tech right yet, but it's growing. We've had 2 6-month periods in a row where we grew 6× in that 6-month period.

Wow. Now, you're on a low base.

Okay. But it means it's 36 times bigger than it was a year ago.

Yeah, it's not going to be small for long. When do we push and start expanding?

But I know we'll drown if I don't get those core workflows tight. We're going to drown in ops.

Yep.

So, we're almost there. We're right about to click in and sort of expand.

Scaling something that's broken is not fun.

Yeah. I like to say, “Let's not scale failure.” That's what I say at the office.

When you're trying to do something disruptive, there's a massive immune reaction that occurs. Sometimes the immune reaction is within your own organization, if people have been doing something else all along, and sometimes, obviously, it's competitors. How do you deal with the resistance that you had—just general resistance?

Yeah. I mean, I like to say that one of the most important things that innovators do is they are really good at finding valuable unknown truths—things that are true and very valuable that nobody else knows.

So, if you create a machine that's finding valuable unknown truths, then you start to know a lot of things other people don't know. And if you know a lot of things other people don't know, you can start doing things that other people can't do.

That's called changemaking. You start to be able to do stuff other people aren't doing, which means you're doing some stuff nobody's seen. So, you start doing a lot of it because you're really good at that core thing, which is finding valuable unknown truths. You start doing a lot of it. You're basically in the changemaking business.

And there is this thing in nature that is everywhere around us. It's called resistance to change. It's in nature. It's natural because nature is a monopoly, and nature abhors change that can cause harm. It's a natural mechanism to slow down progress, to make sure it's the right kind of progress.

And so one of the best ways to deal with resistance to change is to build a massive amount of trust with those you're asking to change, or that you're asking to bring along with that change. You really have to have a philosophy around how to build trust while changemaking, because then you can turn your adversaries into advocates and actually end up moving much faster.

A lot of times, people learn the wrong lessons. They're like, “Oh, man, the Uber thing: just go crack some skulls and kick some ass.” But actually, the lesson to learn is different: if you build trust as you make change, you can actually go even further.

When I knew you at Uber, we were talking about that. That doesn't mean don't crack skulls—you've still got to do that, too.

Okay. Just make sure you know.

We were in the midst of working on an Uber XPRIZE for flying cars. That was part of your initiative, as well as the autonomous car program at Uber. Do you think it was the right thing for those not to continue, or do you think they would have gotten the technology going faster? What are your thoughts? I know it's retroactive.

No, it's okay. It's okay. Look, they killed the autonomous car project we had going on. At the time, we were really only behind Waymo, but probably catching up, and we were going to pass them in short order. So my guess is you can look back at 2020 and say, maybe—I wasn't running the company when that happened—but you could say, “I wish we had an autonomous ride-sharing product right now. That would be great.”

Yeah, I think the same thing here. I think there's still some lack of clarity around flying cars—or, let's say, eVTOLs, electric vertical takeoff and landing vehicles—and how and when exactly it plays out. But it's pretty clear: if you can get from here to there at 150 mph, and it's kind of like a very fast commute, I think it changes the game in a lot of ways.

I think there are companies that are doing some interesting stuff. Joby and Archer are probably two of the more interesting ones, among a number of them. I mean, it's packet-switched humans, where you're moving people through.

Sure.

Well, this is what you just did, right? That's a network for the physical world. You just went into an atoms-based computer, right?

Yeah.

One last question before we go to your questions here, so get them ready. Iterating under pressure. I mean, one of the things that you fabulously did was iterate the company, the models, the products, the services, and so forth. You talk about how important iteration is.

I mean, that's—I guess that's like, how important is breathing?

It's really about how you breathe well. What's your mantra? How hard did you push your teams to iterate, or were you the one driving the iteration?

No, my companies generally run in an empowered way, with checkpoints. We like to say, “Let builders build,” but you need alignment and accountability in order to empower.

Iteration—look, there's a lot to be said there. I'm trying to think of some pithy things to say that make it sound like I figured it out. But you've got to find those valuable unknown truths. You have to have a passion to get it over the line fast and first, and you have to build trust with those you're bringing it to.

Our 3 cultural values at my current company are truth, trust, and passion. One of the things about iteration is really lowering risk. A lot of people say, “I want to have a risk-taking company,” or, “I want to go work for a risk-taking initiative,” or whatever. I'm like, “No, I want to be a not-risk-taking company, or a risk-mitigating company.”

I like to say that innovation equals big P divided by little R. Big P is progress, and little R is risk. Risk comes in the form of time, money, and reputation. So I need to squeeze down risk. If I get it to 0, then innovation goes to infinity. Hold on to that progress while getting risk down.

As you go to the mics, I'm going to ask one more question. You've really been passionate about a customer-centric focus. Speak to that for a moment.

Who's the best of the best of the best at this? It's probably Amazon under Bezos. That was just next level, done so beautifully. It's so clarifying to make a person's happiness or a company's happiness your target. It's visceral. If you start there and work your way through it, you just get clarity.

A lot of people go, “Oh, I'm going to be customer-obsessed. I'm going to lower the price.” Well, if you don't exist in business next year because you went out of business, you weren't customer-obsessed. Customer-obsessed means you have to have a lot of heart for the customer. You have to have a lot of love for them, but you also have to have a lot of ROI.

If you get a lot of ROI with that heart, then you can put in even more heart. I think that's one of the things where the feelings and the numbers come together, if that makes sense. Sometimes people forget that it's about both.

Simon, this is not about technology. It was about your mindset when you had so much trouble pushing the taxi syndicates, and every day it seemed like you had another issue. How did you keep resilience in your mind and your team? What's the mindset you can recommend to us? We are always seeing challenges and changes every day, like you did.

Well, it's interesting because every once in a while I'll get an entrepreneur who comes to me and says, “When do I give up? When do I move on?” I've sort of formulated a 3-step program.

I'd love to know the answer.

I've held on too long a few times. I've never given up early, but I've held on too long.

I have, too.

I don't know if this applies to all areas of life, but it might. I haven't thought about it.

Okay, step 1: Do you still believe? If you don't believe, move on.

Yeah. Obvious.

Step 2: Are you the right person to do it? This goes back to self-awareness and just knowing: What sport are we playing? Am I a support player? Am I the main player? Where do I land? Where am I in this? You need super-hyper-transparent honesty with yourself about it.

Okay, so let's say we check that box, too. Step 3: Am I about to do significant mental or physical damage to myself by continuing? I remember one of my companies. And that's the hardcore entrepreneur way. That's when you're really right on the edge.

I was running a launch company in the late ’80s, way before SpaceX. In the last days of the company, I was making journal entries saying, “The patient’s on life support. Four days left to live.” Yeah, Richard, if you start having apocalyptic dreams or nightmares, you’re probably getting pretty close, in my experience.

Richard, please.

Hi, everybody. Richard Medaf [?]. The question on my mind is the inner journey. You’ve led this industry disruption at scale, and as it got bigger, were there moments when there was some kind of inner story or fear, or when you stopped playing the big enough game? I’m just wondering what those pivotal inner moments as a leader were.

I’m not sure I understand the question.

You were leading change where there was no playbook, right? You were out there at the forefront, writing the rules as you went, with a lot of pressure from various states. I was just wondering what that did for your own personal leadership. Were there moments where you lost your nerve or doubted yourself?

Never happened. Just kidding.

Okay. So, the inner game—what’s the unheard story of what goes on for you?

Look, there’s something to be said for when I finished that company before Uber—the one where I did 4 years with no salary, ran out of money a couple of times, had apocalyptic nightmares, and all this kind of thing. I did a couple of talks, entrepreneur talks and things like this, and one of my aphorisms was, “Money will not buy you happiness, but it will pay for therapy.”

The clearer you are and the more centered you are, the more you can do.

We’ll go to one of our teens. Omar.

Thanks, Peter. Hi, Travis. How do you see CloudKitchens evolving in the near future? What role will AI and automation play in its future growth, and do you foresee CloudKitchens merging with any other AI company to become even more cutting-edge?

Look, the high-level idea of that question is this: Can you get a meal that is prepared and delivered to you so high-quality, so cost-efficient, and so convenient that it approaches the cost of you going to the grocery store? If you do that, you do to the kitchen what Uber did to the car. You turn this thing that we all do for ourselves into a thing that a service does for you.

That doesn’t mean we don’t cook. I like horses, but I don’t ride a horse to work. This kind of infrastructure will be so much healthier. I like to say, “You don’t have to be wealthy to be healthy,” and that’s one of those things we talk about.

What you get in return for this is that you basically get your time back to do all the other things in life you love, and cook when you want. That’s great, but you don’t have to. Similar to what happened with Uber, of course everybody used to drive themselves, and more and more every year, somebody else is doing it for us, or there’s a service that does it for us, and ultimately an automated one.

As it relates to AI, I think everybody is very familiar with ChatGPT, DeepSeek, Grok, this kind of thing. I call that Bits AI. That’s AI for bits.

There’s a whole other thing that we’re probably going to start seeing more of, which is what I call Atoms AI: AI around the physical world. You could say the first iteration of that is what you see with the Waymos rolling around, right, and autonomous cars generally. But what about humanoids? How does a machine move through the physical world and act in the physical world?

There’s a whole other set of models that are going to need to be invented to make that real, sometimes borrowing from the other side, but it’s a different ball game.

Thank you. Gustavo on Zoom. What’s your question, Gustavo?

Hi, Travis. I’m from Brazil. I am a serial entrepreneur. I have a $1 billion company in my track record that I co-founded. In my experience, luck plays a bigger role than my vision and ability. In your view, does luck significantly contribute to the success of a business? Could you venture a percentage? If so, how important is luck, Travis?

Okay, so I like to play back. Does anybody here like to play blackjack? Anybody? All right, there we go. In any given game, luck plays a real role, sure. But if you play 100 games, it doesn’t. That’s a really good analogy.

The way to think about it is: Is the success that you have due to 10,000 decisions, or is it due to 2? If it’s due to just a couple, then you’re lucky. If it’s due to 10,000, you’re not.

Love that. Cartique [?].

Thanks, Peter. Travis, thrilled to have you here. My question is actually—and I’m going to reference George’s question up there, if you don’t mind—about what framework you have and are deploying to assess timing, but also valuable unknown truths.

Yeah. I mean, timing—you get burned. A lot of the other questions I get are, “What did you learn at the last thing?” or, “What did you learn?” And I’m like, “Well, I learned all the things I shouldn’t do.”

All of that is really about seeing the future, right? A valuable unknown truth means you see something nobody else does. It means you believe something that is against what everybody else believes. It means you have to see the future.

You have to see the difference between perception and reality. Everybody thinks reality is here—that’s their perception. They think reality is here, but it’s actually over here. The distance between the 2 is the innovator’s playground.

You have to get good at seeing what the future really is, being very skeptical of everything coming in, analyzing in a very truthful way, and getting good at it. What is actual reality, not what people think reality is? Look for those moments when what people think is very, very different from what is real. That’s where all the good stuff is.

Beautiful. Give it up for that. John.

Yes, me. Hi, Travis. Thanks for being here. I have a really important question. How has your leadership changed since everything that happened at Uber?

Yeah, great question. Before Uber, the largest company that I had run was a 12-person company. Everything I’m saying right now comes from a journey where you learn every day and deeply commit to getting better every day—from 12 people, running scrappy, not knowing what the hell you’re doing, and not paying yourself for 4 years, to thousands.

How big was Uber at that point?

It was like 15,000–20,000 people.

Yeah.

And then several million drivers. It was all—so what I learned is that everything I’m saying, you iterate on all those things as you go.

We’ll go to Elin. Elin.

Thank you, Peter. What’s your advice on an industry that’s super-segmented, such as construction materials? I’ve been trying to digitize that industry, and I’ve been wrong for the last 5 years.

Look, we buy property and do construction. That’s part of what we do. I’ve sort of dabbled in modular construction. The promise of it is so beautiful. Of course, the issue is local contractors and subcontractors.

We did this thing where we would manufacture these beautiful kitchen facilities off-site, just really epically, just wonderful. But then it turns out, okay, you have to put them together like Legos on-site. Now I have to train a set of contractors to build it.

In the first go, we had 1 that could do it. He wakes up one morning and he’s like, “You know, I don’t feel like doing it today.” And you’re like, “Please, pretty please. We signed a contract. You’re supposed to do it.” Then he’s like, “You know, I think if you paid me half a million dollars more, I would do it.”

So you get into this world where it’s against the contract you signed, but they’ve got you. Then you’re like, “Okay, now I have to train all these contractors to do this thing so that I can manufacture somewhere else and then put the pieces together.”

My point is that there’s a lot of change that needs to happen. It may be that we only get there when we get into true robotic construction on-site, or I don’t know. I haven’t solved this.

Vertically integrate everything again.

Yeah, I don’t know. I don’t have the answer for this one, but I have so much hope.

Thank you. Yeah, Jacob, one of our teens again.

Hello, Travis. Do you believe that the mass disruption of currently existing industries happens based on factors within or outside of your control when you’re creating a business, based on your experiences at Uber?

Look, nobody’s in control of the world. You have to see where the world is going and build for that world. Sometimes the world changes in ways that you didn’t expect, and you have to adapt to that.

We’re only in control of what we do each day, but most definitely, most of what’s going on we’re not in control of. I think that’s a super-important thing for people to realize, because a lot of people get really attached to this idea of controlling stuff, and they get into weird spots when they do that.

Now, there’s the other side of it. You get so used to not being in control. As an entrepreneur, you wake up and you’re like, “What crazy, effed-up thing is happening today?” And you just go, “Yeah, I guess that’s what’s happening today. Let’s go.” But then you almost lose your spine. You almost become an invertebrate because you’re like this blob that’s just cool with whatever’s happening. So you have to sort of keep the fight to try to make the world the way you want it, but you can’t get attached to trying to control it. It’s like this balance. It’s amazing.

Varun on Zoom, please.

Thank you, Peter. Hi, Travis. I’m dialing in from Dubai. My question is: The global consulting industry is highly fragmented, and as we live longer, there are going to be a lot more people who work in the industry who will become consultants. It’s really fragmented, and it’s a large industry. Is it a good idea to consolidate this industry using the Uber model?

Push a button, get a consultant. I was just kidding. People are like, “Okay, yeah.” No, I think the consulting world is so much about what can be done with AI.

Yeah, I think consulting is about to get radically transformed with AI right now. If you’re sort of a traditional consultant and you’re doing the thing, executing the thing, you’re probably in some big trouble. If you are the consultant that puts the things together that replaces the consultant, maybe you’ve got some stuff. You’re basically going to those, as I previously mentioned, profitable companies with competitive moats, making their moat bigger and their profit bigger. That’s probably pretty interesting from a financial point of view.

I think consulting is a scarcity-mindset business. You build a wall around these consultants and meter them out a little bit at a time, and that will get completely disintermediated by AI.

Tad, let’s go. You can push a button now and get a consultant. It’s called Deep Research. We’ve been talking about it for the last 2 days. It’s pretty cool. Let’s go.

Travis, what advice do you have for investors and board members who want to make entrepreneurs like you successful, or who are working with people like you? What are the tailwinds and headwinds that they create?

I’m going to say some provocative things.

I was hoping you would.

All right, let’s go. I think there are a lot of investors who make the pitch of being founder-friendly. They make the pitch as if their job is to serve the founder, but that’s not their job. Their job is called capital allocation. Once you make that leap—that you understand that your job is capital allocation and not the happiness of a founder—then I think a lot of things actually get into a much better spot.

Generally, from a founder’s perspective, the way a founder should look at an investor is not, “Which investor is going to be founder-friendly or help me?” I’ve just never seen that work. It’s which investor does the least amount of harm.

The way I look at it is: Okay, you’re playing a chess match. You’re like a grandmaster chess player, right? You’re playing another grandmaster chess player, and you’re playing this game, this match, 60 hours a week. Sixty hours a week is what you’re doing, and you’re pretty exhausted from doing that, but you’re passionate about it. Then there’s this chess enthusiast who doesn’t actually play chess trying to tell you what moves to make on the chessboard. You’re like, “Homie, let me tell you the 18 reasons why that’s not a thing.” It’s not their fault. They’re not playing chess. It’s okay. But the do-no-harm is a real thing.

Look, governance matters. But a lot of times I think investors get in the mode of thinking they’re the idea guy. If the investor is the idea guy, you have the wrong company. Getting a board that supports you as a moonshot entrepreneur, that gives you the freedom to do riskier things—it’s got to be smart risks.

Yeah, right. Of course.

But still, you, as a board member, should be like, “Are you smoking something?” The response should be very good, but you should be asking, “Are you smoking something?” because things can get weird.

Thank you. This is such a wonderful conversation. We were having dinner last night. It was a group of founders, and we started to brainstorm around what’s happening over the next 5 years and 10 years. Will robots and AI replace all of our jobs? How do we prepare for that future? I’d love to get your thoughts because it’s coming up fast.

This is a hard one to answer because the history of technology basically says we’re fine. But a lot of times history says you’re fine, and then things change. So it’s like, I don’t know. The one thing we’ve got going for us—this is a weird one—is that robots don’t have bank accounts yet. We had a 2-hour session on crypto last night.

Easy, easy. Okay, hold on. Let me just roll with this for a second.

Okay, let’s just assume robots don’t have bank accounts. Why does that matter? Because if you’re making something cheaper, it’s cheaper for a human. It means that they have more money to spend on other things. When they spend it on something that’s powered by AI, it doesn’t go to a robot’s bank account. When some founder makes a ton of money doing whatever and they’re going to spend that money, it doesn’t go into a robot’s bank account. So, in a weird way—I know this is a little bit niche with the way I’m answering this—I’m more optimistic than pessimistic.

I’d say the other thing that I’d put out there, which I think might be more generally applicable or make more sense, to be honest, is that I’ve done companies that have failed. The main thing when the company failed that I experienced, or that I saw, was that nobody wanted the thing that I made. If somebody wants to go and make a thing that doesn’t make people’s lives better, they should count on failing. If you want it to succeed, you better be making somebody’s life better, and probably a lot of people’s lives better, if you want it to succeed.

That’s where things ultimately go: The machines, by their nature, their DNA, are serving us. Just throwing it out there.

Time for one last question. Rachel, thank you.

And Travis, Uber changed my grandparents’ lives, so thank you. My name is Rachel Odin, and I am an Ascend fellow and also a new founder. I am building an AI-driven company to revolutionize how we form, sustain, and grow our most intimate relationships.

The question I have for you is that Uber wasn’t just a company. It introduced an entirely new market. It was disruptive innovation. So how did you balance needing to educate people on the market with building your product, refining your business model, and scaling?

Well, you hear about somebody’s first experience with Uber, and the fortunate part is the product itself is the education. They’re sitting at a dinner table, and you’re like, “Check this out.” It was naturally social. A large percentage of the app installs were at restaurants, right? A large percentage of the app installs were due to one person introducing the thing with a link: “I get a free ride when I give you a free ride.” We called it Give/Get. It was like a third of all of our new users up until 2015 or 2016. So the product itself was the education.

I think it’s pretty weird that one citizen taking another citizen across town became somehow really controversial. It’s kind of weird that that was controversial, but it was. So we had to sort of deal with the controversy as we went. Maybe it gets back into that changemaking thing and trust, or lack thereof.