transcribe

Tony Xu on AI Consumption, Atoms vs. Bits, and the 1% Better Mentality | Ep. 55

Uncapped with Jack Altman · 47m · transcribed Jul 2026
More from Uncapped with Jack Altman Business
𝕏 Share ▶ YouTube 📥 PDF 🤖 .md

Section Insights

# 0:00

The Battle for Attention and Resources

What is the significance of AI spending in companies like Uber?

The discussion highlights the challenges companies face in effectively utilizing AI technology to solve real-world problems and improve efficiency. The speaker emphasizes that technology should ultimately aim to make things cheaper and more effective, but the rapid consumption of resources, as seen with Uber's AI budget, raises questions about the tangible benefits gained.

  • Companies are grappling with how to effectively leverage AI technology.
  • The ultimate goal of technology is to solve problems and reduce costs.
  • Resource constraints can complicate the implementation of AI solutions.
# 9:29

The Importance of Physical World Interactions

Why is it essential for technology to engage with the physical world?

The conversation underscores the need for technology to extend beyond digital interactions and facilitate real-world actions, such as moving items or providing services. The speaker argues that software should ultimately serve physical needs, emphasizing the importance of tangible outcomes in technology's evolution.

  • Technology must bridge the gap between digital and physical interactions.
  • Real-world applications of technology are crucial for improving lives.
  • The future of personal assistants relies on their ability to perform real tasks.
# 18:59

Trends in Consumer Spending and Restaurant Growth

How have consumer spending habits changed over the decades?

The speaker discusses the shift in consumer spending from groceries to restaurants over the past 75 years, highlighting the increasing preference for food prepared by others. This trend correlates with the rise of dual-income households, indicating a broader societal change in how people value their time and convenience.

  • Consumer spending has increasingly favored dining out over cooking at home.
  • The rise of dual-income households has influenced spending habits.
  • The restaurant industry has shown consistent growth despite economic fluctuations.
# 28:29

Challenges in the Advertising Business

What are the key challenges in balancing advertising effectiveness and consumer experience?

The speaker reflects on the complexities of the advertising business, noting the difficulty in creating ads that yield high returns for advertisers while also ensuring a positive experience for consumers. The inherent conflict between these objectives makes it challenging to maintain a balance that satisfies both parties.

  • Creating effective advertising that also respects consumer experience is challenging.
  • There is a constant trade-off between advertiser returns and consumer relevance.
  • Successful advertising strategies require careful management of conflicting interests.
# 37:59

Growth Strategies for a Mature Business

What strategies are considered for business growth in a mature market?

The discussion focuses on identifying key levers for growth in a business that has already achieved significant scale. The speaker emphasizes the importance of solving existing customer problems and anticipating future needs, suggesting that innovation and adaptation are crucial for continued success.

  • Identifying and solving customer problems is essential for growth.
  • Even mature businesses must innovate to remain competitive.
  • Understanding market dynamics is crucial for strategic planning.

Transcript

0:00 There are battles for attention, you know, bits and, you know, battles for what's happening in the physical world, you know, for the atoms and and and that we largely occupy ourselves in the second category, right? We are in the war for atoms because, you know, one day ultimately like what's the point of having a personal assistant if it can't actually do real things. >> Tony, thanks so much for doing this. One of the things I wanted to start start with was this concept that if you'll indulge me on it because it's something I've been thinking about a bunch kind of around the idea of like AI spend at companies. And I think the the trigger sort of anecdote that I saw that I'd like to get your take on was Uber was basically saying, you know, in the first few months of the year they blew through their whole AI budget. They were just, you know, rampantly consuming tokens and what did they get out of it?

0:52 Did they get more rides? Did they get more drivers? did anything you know their margins improve like what happened and I think they were kind of saying like it wasn't obvious and you're obviously you know the leader of a company that's super metricsoriented and has executed super well so I I assume you must think about this but I'm just curious kind of at a high level how you think about like what you know what is a large rising line item and you know as as it continues to rise what do you what do you care about? Yeah, I I I I think every company right now is trying to figure that out where, you know, the ultimate goal of any technology is actually to hopefully solve a problem and actually make things cheaper. I mean, that's the that's the goal of technology, but as you know, especially when you have constrained resources like compute in the case of these large LLM companies, you know, sometimes you don't get the timing of these things always exactly when you can deliver the outcomes. or you don't get the cost profiles or the efficiency profiles to serve the technology to match exactly when you can deliver the customer outcomes and I think that's the world in which we live today. So I think every company is trying to figure this out in terms of how we're thinking about it.

2:05 you know first and foremost is like what customer jobs can we actually solve and so I I do think like when you have new technologies there's always going to be this period of inefficiency and you know almost like discovery >> where you got this new toy you don't know exactly what you need it for. you probably kind of know that you don't really need a you know frontier model to know what the weather is perhaps but on the flip side you don't also know the limits or the ceiling of what the technology can do and you kind of want to know that and and so the way at least I think about how you do this somewhat efficiently even though by definition you're going to accept inefficiency in the discovery process or in the invention process you want to do it in the most contained set of ways that deliver customer outcomes. So you actually want to put customer outcomes and start there and then actually give teams, you know, as many shots on gold towards those outcomes as possible.

3:07 Doesn't mean you get there, but at least it's directed. Yeah. >> And it's intentional and it's not, you know, entirely just yolo. there's obviously some of that. but I think if it can be directed towards you know in our case consumers, merchants, dashers, we've actually found some success. >> Well, it's funny cuz like as a software company which is like where so much of you know the AI productivity is happening right now. It's actually harder in some ways for them to measure the like end outcome of the work that their engineers have done than maybe yours where you can say hey I can measure did we complete you know more deliveries? Did we get more restaurants?

3:44 Do we have new users? or you know all the things that you're tracking. You know I was sort of like u I was reflecting that over the last few years you've kind of like watched this egg move through the like snake in the pipeline from you know first there's the buildout and there's chips and then you get data centers and then it's like oh man are people ever going to use this and now they are using it and the token spend is ramping but now it's like we've got to turn the token spend into like burritos at people's homes and like you're in like as good of a position to see that as anybody. So like are you are you like you know we're going to have certain teams try to move the metrics in a market with AI or is it kind of just like bottoms up let the teams explore whatever they want with AI?

4:22 >> Yeah I think yeah it depends on the team. So if your product teams you you're exactly right you can direct teams I think towards metrics that matter to customers. If your merchants on Door Dash today, you know, they're onboarding 35 to 50% faster, because we can actually, you know, have AI produce all of their catalog or their menu in the case of a restaurant. you know, help edit your photos, help set up, you know, what might be the best way to describe yourself. especially if you're a new retailer, coming online in into a city. so there's metrics like that that are very trackable, measurable, have immediate positive benefit to in customers. similarly with dashers, you know, you we we've had AI help detect, you know, both fraudulent fraud fraud as well as safety incidences before they occur and you can stop those and take preventative action if you knew that was actually happening. So there's things like that where you can you know get immediate benefit you know so immediate customer benefit and but but you you said something earlier in your question you know what about like software teams for example and I think one of the things you notice there is that only certain parts of a software engineers day is writing code and it's great that like we have models today to write the code for us but that helps maybe the whatever 25 30 50% of our time that's actually you know shipping code but what about everything else you know that has dependencies in product reviews design meetings you know alignment with business teams etc etc etc that also has to change if that doesn't change and come together you're not going to be able to just have perhaps the productivity gain that you hope to have and I think that's what you know companies like ourselves but I'm sure a bunch of companies across the industry are trying to figure out and getting those workflows right so that you're not just AI native from code development but you're AI native in how you actually operate.

6:29 >> Have you done anything to like track different engineering and product teams inside the company comparing like how much more productive they become with AI usage like ones that are using it more and less heavily and things like that. >> Yeah, I mean you get all these facts and I'm sure you know as with all things there's always a distribution of outcomes. you know, maybe the average is 50% more productive. but you know, you have engineers who might be 35 times more productive. It's crazy, >> right? And and and it is crazy. But and and and it's always fun to study that distribution, but and but again, like I I think that you know what's the value of that? The value of that is really knowing what the art of the possible is, especially in understanding your workflows. But what you know someone like I have to do is I have to think about well how do you create that working environment for everyone you know because it's it's great that we have you know engineers who are 50 times more productive how do you actually get everyone but is that an observation to that new ceiling >> exactly and the question is is like is that an observation and somebody's going to be far out or is there a learning there that you can go replicate it >> yeah exactly and and and and I I believe it's the latter because with all things especially if they're newer you're always going to get a distribution of outcomes Right? There's always, you know, in sports, in academics, like there's always a distribution of outcomes, but that doesn't mean there aren't classes and teams and, you know, good things that you can teach as a coach or as an instructor so that you can get the rest of the class, you know, to move to the higher plane.

7:59 >> Yeah. You know, for a lot of companies when AI came around, it was either like the most like, you know, kind of petrifying new technology ever or it was like the oh my god, thank you Tailwind. And I feel like for Door Dash, it was probably like neither. And you tell me if you object to this, but like I would say like, you know, you're mostly operating in the world of atoms, not bits. you know, you've got this two-sided and three, you know, threepart marketplace with, you know, restaurants and dashers and consumers and all these things and, you know, you're very, you know, related to cities and all of that stuff. And so, I would think it's both insulated from AI in a good way. And then you also maybe didn't, you know, it's not like your business started growing 400% faster because this AI thing showed up and, you know, wasn't there yesterday. So, like, how is that been? Is that like is that right? Is that like about what it's been for you?

8:48 Has it been kind of just insulated and that's both good and bad? Yeah, I mean we're not selling, you know, burritos or Nike shoes or groceries as fast as tokens, you know, at some other companies, but yeah, look, I mean, I I think you're largely correct. I remember I don't know, I think this was 2021. This is Yeah, this is 21 before the arrival of Chat GPT. We were playing around with some of these models. I mean, I think it was like >> it was like GPT 2.

9:20 something like that, right? >> And just kind of getting a sense of what the world could look like. and obviously, you know, the models back then are very, very different from the models of today. but even then we kind of had this description of the world where there are you know battles for attention you know bits and and and and and you know battles for what's happening in the physical world you know for for the atoms and and and that we largely occupy ourselves in the second category right we are in the war for atoms and moving things around and if we can do that and we can build a catalog for example for where every item exists inside of a city or every parking spot exists or all of this information. We can as the kind of company's battle for attention start maturing that we can really partner and work together in very productive ways because you know one day ultimately like what's the point of having a personal assistant if it can't actually do real things. Right. Exactly. Right.

10:24 And so and and that's kind of how we thought about it four or five years ago. It's actually funny on that point like you know I it it strikes me that like one of the it's very cool but you know one of the things that I think there's room for improvement is that so much of tech is just like typing on our screens and we're all living in the computer but at some point to make lives better you need to you know move stuff around. You need education. You need stuff that happens you know in a hospital. Exactly.

10:50 >> Yeah. You know and so it's you need you need you know real estate to be developed. It's like all this kind of stuff that's but it's all physical at the end. >> Yes. And so in some ways it's like all the software ultimately does need to be in service of stuff. >> Yes. >> Yeah. >> Yes. I mean and and this is and and this is exactly why you know what is I guess it would have been 5 years ago at this point. the the kind of the the strategy was very much no play the game that you were meant to play. The game we're meant to play is the game of atoms and you know be bestin-class at that.

11:20 And not only is that good for our business, but I also think it's great for our relationship over time with these companies as they, you know, finally build out, you know, some of their assistance, and build out super intelligence and build out agents that can actually do things versus, you know, what they currently do. >> because we naturally will need one another in order to be useful to the communities that we serve. Has there been any impact on like recruiting or things like that when everybody's like losing their mind on AI and you're in this different thing that you're like I know this is really important and it's going to become even more important eventually but like have you had to work through any of that?

11:57 >> Yeah, I mean it it's definitely a battle. We are an applied AI company and we have been you know pretty much since day one of the company whether you want to call it math or machine learning or now LLMs or you know we can use whatever terminology we want but as the years have gone by we are an applied technology company that that also solves real world problems. So yes, it's it's certainly difficult when you're competing in places, you know, where activity is really hot, you know, in machine learning in folks who are looking to do applied research, it's very competitive right now.

12:31 >> Yep. Are there things like as you think about, you know, like your next frontier and where you want to go and you look at the AI stuff, it does seem like some of it is pretty direct like you know, for example, I could imagine like robotics maybe could matter at some point. I could see self-driving matters at some point. Like what are the like next parts of AI outside of like this like the LLM on the software side that you care about?

12:53 >> Yeah. Well, well, I mean I would say first, you know, before we step away from, you know, some of the LLMs, we do care about, you know, what LLMs can do and and I think they have gotten materially smarter and more powerful. not not just you know helping with coding which may be the most prevalent use case today and certainly in the most prevalent spend consumption. >> Well it's definitely a good input to you for building stuff.

13:17 >> 100% 100% but but even you know for instance you know one of the things that you know we always ask ourselves is like how can a technology actually improve outcomes for customers. so if you take as an example the Door Dash app, you know, we've grown a tremendous amount over the last even 5 years, forget like, you know, the 13 years we've been doing this, just in the last 5 years, kind of the same arc in in which, you know, some of these LLMs have grown up, you know, Door Dash has moved from one product category, restaurants, one market, US into, I mean, virtually every retail category, the leader now in, you know, deliveries of grocery, convenience items, alcohol items across 40 countries.

14:02 >> That's there's lots of positives with that. One of the challenges for a consumer though is boy, it's a lot harder to use an app like Door Dash because you now have restaurant things in there. You now have grocery items in there. You have retail items in there. You now have the ability to make reservations or get deals inside of restaurants. It's getting more complicated to actually use our product. LLM, you know, really can help with that. And you know, just as I think there will be, you know, personal agents soon, the Door Dash app should be a personal agent. It should be a personal agent to help you do anything inside of your city, >> right? I can't think of any other product and any other utility, frankly, greater than the number of connections that you can have between you and the businesses inside of your city. And so those are the kinds of things that you'll see us launch just with LLM, right? you should be able to have an easier time actually doing research on what items you want to buy or what might be great recommendations for you based on all the history that we have across tens of billions of orders on you >> and you should pretty much be able to get anything to your house and not have to shop for it.

15:10 >> Yeah. Or have the choice of getting it you later when you're home or because you may be away or the next day because it's better for you. you're exactly right and so I do think that there's still a lot of excitement just in the LLMs >> but you're right if you want to go beyond that >> there there there is a a lot of work around the physical space so for example we've been working on autonomous vehicles since 2019 >> and a lot of that started with kind of traditional traditional systems kind of like you know with machine learning yet traditional ways of thinking about how you know, use those techniques to build things like recommendation systems. LLMs kind of put a complete new spin on it and didn't really require any of those techniques. something similar is happening in the physical world where it used to be if you wanted to perhaps you know drive autonomously you know inside of a market a lot of what you would do is actually build mapping systems and and almost like heruristics and rules to create an engine in which you can make good decisions and and kind of weigh them in real time or perhaps you can you know put all of this into a neural net and and take similar techniques that some of the LLM companies are using and actually make, you know, even faster and better decisions, right? And so there there are things like that that are going on that allow us to do some of the work that we're doing with autonomous delivery, you know, faster as an example.

16:43 >> Yeah, totally. This is now like going way back in time, but are you surprised how much people are willing to consume on apps? You know, obviously when you started the company in 2013, this was the plan, but like has it gone like a lot further than you expected? You know, like people really will just like order, you know, a burrito for, you know, a lot of money, you know, and it probably like and it's worth it like turned out that it's worth it to people, but like was that obvious to you in 2013 that it would go this way?

17:13 >> No, it was not obvious. >> People pay like 35 bucks for a burrito. For example, when we launched our first partnership with a a national brand that was July of 2015, I believe with Taco Bell. I was actually quite skeptical whether or not it would work. You know, on the one hand, you know, you have this legendary brand that's never offered delivery before being offered for the first time. That was the bull case version. On the flip side, you know, you're right. You're going to pay a premium, you know, to get that order.

17:49 Yeah. >> And and and >> which wouldn't have been intuitive because you said, "Well, I can get the, you know, I can I can get this order for $4 in store." >> Yeah. Exactly. And and and so that was not obvious to me in 2015 of what would happen, the the bull or the bear, you know, version of that. But clearly people love getting Taco Bell delivered. Of course. >> Well, that's also I mean it now makes sense to me in a weird way cuz it's like you know you you save a lot of time going back and forth. You can use that time to do all this other stuff. You can use that time to work like you know and I think more people are familiar with those kind of calculations now and all that stuff.

18:20 >> Yeah. I think it's that I think I I think there's a lot of things Jack. I mean like one of the things that I remember even in 2013 looking at as just this marvelous like fact that that that only goes in one direction. There there's a few of them. You know, one of them is that if you looked at if you looked at just food consumption, you know, in the 1950s when the US government used to measure this or when they first started measuring this, something like 70 to 80 cents on the dollar was spent on grocery. Okay, this is like in the 1950s. If you looked at 2013 when Door Dash was founded, that number was getting closer to 55 cents towards groceries, you know, 45 cents towards restaurants. Today, it's closer to 55 cents towards restaurants, 45 cents towards groceries. And so over a 75 76 year, you know, arc, yes, there's ups and downs, but if you look at the trend line, it kind of goes in one direction, which is in the direction of, you know, >> food prepared by somebody else. Do you know by chance if like the relative cost of a steak you made through your own groceries versus a steak prepared by somebody else if that ratio of cost has ch like has one gotten more expensive relative to the other?

19:36 >> It's a it's a great question but I think it depends a lot on how you value your time. So this gets me to another fact that I think is pretty interesting which is if you looked at the percentage of dual income households same time period 1950 to 2020. So this you know this goes way beyond AI or COVID or any of this sort of stuff. you find that the percentage of household has gone from like a quarter you know dual income to almost like 70% right you know present day. If you look at the number of restaurants as another example just just total number of restaurants total count okay in the 60 to 70 years in which this is measured there's maybe two years in which the total number of restaurants in the current year doesn't exceed the previous year. So even if you know new restaurants you know which is a very tough business as you know I mean having worked as a dishwasher my mom's place I definitely know how hard it is yes they there's a risk of going out of business >> but there's always more than enough that replenishes the bucket every successive year. I think when you put together some of these, you know, facts over like 70 plus years or something, they kind of spell out that >> people whether it's through, you know, dollars or time >> are expressing the fact with their activity, not just their words, but their activity that they must value if somebody else made them mistake.

20:56 >> It's interesting because like, you know, the the stat about, you know, dual income households, it's like that that's true, but also like the you know, the cost of a home has gone up crazily. I think you people probably feel like it's harder to get like the home and the car and everything than they used to and all of that. Yes. >> So, it's a little counterintuitive that people's willingness to spend on food has done what it's done.

21:21 >> Well, I think a couple things. So, you know, the first thing I would say is, you know, food happens 20 to 25 times a week. It it's not like buying a house. >> Yeah, that's right. >> So, that's that's the first point I'd make. So, so even if you're the most avid cook, right, that you love making, you know, food, >> what's 20 minutes? >> Very difficult. It's very, very difficult to do it 20 to 25 times a week. That's the first point. I think the second point, and I think you make a very good point on the affordability challenge that we have as a country and probably globally actually, not just here in the United States. is that people always want to find a way to feel good.

21:58 >> Mhm. >> Okay. And and and and I think you know purchases especially on food you you know they don't just become like an indulgence anymore. They become like you know the ability to actually you know feel good. >> It feels great to hit a button and a pizza shows up. It's it's great. Yeah. >> To to feel good more than once and also do something that you know you need to do. Y >> food consumption.

22:20 >> That's right. How much do you think about like consumer trends in general outside of food? Like do you need to just be myopically focused on food or is it worth your time and headsp space to care how people are spending their energy on social media or what's going on with Kouchy and Poly Market or other consumer like does that matter to you >> to understand >> as as a consumer business with hundreds of millions of customers now absolutely you have to think way beyond food and as a company that frankly doesn't just do food anymore increasingly our order orders are coming outside of food we have to pay attention >> so what are like some of the other consumer trends trends that are like not about food but that are important and interesting to you.

23:01 >> Well, I think you named one of them which is this affordability piece which isn't just about food. People are looking for a affordability across every segment. >> Yes. housing, transportation, eating, groceries, healthcare, >> I education, I mean we can keep going, banking, I every every category I would say affordability is a huge deal. >> Y, >> huge premium. And so a lot of you know, what we're thinking about is how do you continuously do two things? One, continuously bring down costs and two, how do you bring more value? and and and I I think those are very hard to do things. we we we again like mainly try to stay focused in the world of atoms though because one of the things that I think people on the software side perhaps don't appreciate is you know all this information that you can get in software kind of sometimes gives you this perception that you can structure information pretty easily and maybe control information and experiences pretty easily end to end.

24:07 That is the complete opposite in the game of atoms where everything is an edge case. Every day there's this thing called traffic and weather. And every day by definition it's not perfectly predictable. And we can argue it's rangebounded or not. But look, if there happens to be a traffic jam and you know something takes 20 minutes longer, that's a that's a real problem for that one customer. A lot of what we're doing is actually just staying as expert and proficient as we can in that game.

24:38 >> What's cool about it is by being so good at logistics and costs and all and coordination and all of that, it's obviously going to apply to stuff outside of food. but even even just within food, it seems like, you know, to bring more value to the customer. I mean the obvious ways you could just keep chipping away at you know the cost which I assume over time you ought to be able to get to a extremely low place I would think >> with food but also with inventory of other types of products right for instance you know one of the challenges in you know grocery or retail is >> well there's actually separate challenges in the in the case of grocery most grocerers don't know what items are on shelves and it's not because of bad technology or outdated systems or a lot of different systems. There are those challenges, but it's honestly also structurally because consumers who go inside the store move things around, right?

25:30 >> or CPG companies, you know, want, you know, certain items to be promoted or not promoted and those things change quite often. And as a result of that, you know, that becomes really messy, right? So, how do you get great at that? You know why? Because if you don't get great at that and you make mistakes or you have to make a bunch of substitutions, that's extra costs. backs to your point around affordability. That's that's cost. Even though it's not just about the price of the items, but it's about everything surrounding it to support fulfillment. In the case of retail, if you didn't know that, you know, the pair of shoes that you wanted might be a half size off, you know, because it doesn't fit great, that's extra cost that's going to get returned somehow or refunded, you know.

26:10 and those are all of the challenges that we try to obsess about. What are the like tempting adjacent things that kind of make sense for you to do that you've like said no to in the name of focus? Like you know as an example as I was just listening to that like you know do businesses ever want to use you to like you know work with their own suppliers or things like that and then have you said you know what that's just too far a field from what we do like are there are there close by things that you're constantly saying that's a good idea but it's not a great idea we're just not going to do that.

26:38 >> Yeah it's a good question. I mean a lot of the separation between good and great internally right now is around sequencing you know and and and so you know for instance I had no idea back to one of your earlier questions about well how big could food be >> it's very hard as an entrepreneur you know when you're working out of your apartment to know >> then that you know what that size could be one day >> turns out a lot of people eat food >> turns out a lot of people eat food it also turns out it's a lot harder than we and we worked on it for seven years before we moved to category number two which was groceries. It's funny. It's like I could see you getting started like a this could be a $5 billion company and now you're like this could be a $500 billion.

27:20 >> Actually in our Y cominator application there's a question I don't know if they ask it anymore which is how much revenue do you think this company can make one day and I remember we were just operating in Palo Alto at the time and so I counted up how many Palo Altos there could be and how many orders we could do in each one of those types of cities and I estimated something like hund00 million of revenue or something like that. Thankfully, we're a few orders of magnitude off. That's funny.

27:45 >> But but that's true. That's a But but but to your point about >> No, but it's surprising. >> It's very surprising. And so I think a lot of times it's you you kind of as an entrepreneur have to take the greedy algorithm, right? You have to keep going all the way still on a winning swing. Why? Why would you get and that's kind of how we think about it when you ask, you know, about like, oh, are there other things we could do? Yeah. The question we ask is >> yes. So then your job is to basically just say >> and what's the opportunity cost? What's what's the opportunity cost relative to the current thing? But you know, I mean, but but Door Dash is a company that has built a lot of different things now. You know, we have obviously a business in restaurant delivery. We have a business outside of restaurant delivery, a business outside of the US, a business in advertising, a B2B business where we basically take everything we've built for ourselves and we give it to merchants so they can stand up their own first party channels, right? And so that came from somewhere, you know, but a a lot of it is just constantly weighing, you know, is it the right time now versus is it, you know, a bad idea or a good idea.

28:50 >> How hard was the advertising business? Feel like those people >> hardly. >> Yeah. Well, I I I I think just building ads, not not that hard. building best-in-class returns for advertisers while you also have best-in-class returns for the consumer experience. Hard because fundamentally there is going to be a conflict where if you have an advertisement show up in a product that may not be relevant to a consumer, you have a challenge where you're taking some sort of a trade-off. We can say that oh it's worth it here and there but you know before you know it you know after a few years or a decade >> it's a slight tax. It's very very difficult to kind of unwind all of that especially given some of the attractive profile of the economics behind advertising. And so I think well one of the things I'm really proud of you know I think the team gets a ton of credit for being the fastest company in history to hit a billion dollars in ad revenue but I'm more proud of how they did it because it's constantly constantly >> I mean you know fighting the restraint effectively or or living with the constraint that we must achieve both objectives >> best-in-class returns for advertisers as well as consumers. I mean, you know, it's like fa I think like probably the Facebook ads team, the early ads team must be like one of the greatest.

30:08 It seems like in in some ways it actually seems to me and I obviously I know you're on the board there at at Meta, but I feel like in some ways it seems like those cultures are your culture and the metaculture probably have a lot in common from at least from the outside in terms of just like you know extremely metrics driven, a lot of testing and trying things like very focused on like you know results and stuff like that. Is that is that is that like accurate? Is that like what like was the ads team like an even more distilled version of all of those things?

30:35 >> I mean I think you're right in that in saying that you know the ads teams at some of these largest tech companies in the world are some of the you know candidly most impressive teams because you know it's carried them so far. You know, I think we forget that some of these companies and products that you're we're talking about here are more than two decades old now at this point. >> And not only do they have the reach of billions of users and things like that, if you looked at some of the, you know, latest results from these companies, it's incredible the the the business growth.

31:09 >> I mean, also you've seen you talk to people from, you know, Google or Meta Ads. I mean, it's brilliant people. Like, it's very hard. >> Yeah. It's and and and and and the constant working that problem, right? The the and it's and and you're right that does share, you know, in some ways with the Door Dash side where maybe we work in a different space. It's not fully in our control. It's not all about, you know, digits and and and and attention. It's more about the physical world. And we have a lot more constraints where we have to kind of take what the the def you know what the defense gives us so to speak where you kind of are taking what's happening in the physical world and then you have to react very quickly to it because we don't get to control sources of demand or supply really. but it's similar in that you you have to be very objective and unemotional about what is best for customers while living within constraints and then getting 1% better every single day and not taking for granted that you can't because I think sometimes it's easy to say especially intellectually speaking that like oh we've solved the problem >> finished you know like delivery is finished >> and I think if we ever thought like that I don't think Door Dash can continue growing. And I think that we've continued to our teams have continued to just do better by increasing our selection, making fees more affordable, increasing the quality and reliability of our network and delivery, improving our customer support constantly, every single day 1% better.

32:39 >> and that's that that that leads maybe not in one sitting, but over an accumulated period of time, a lot of benefit and surplus. >> Yeah. when I mean when you started the company obviously you were not the only or first person doing this and there's probably you know a lot of things that people could say about you know what led to all the success and there were good decisions you know you starting in the suburbs I think was like a big one and probably the way that you you know went about acquiring restaurants and all these different things but in some sense also I think it's probably like a fair kind of like summary of it all and what you guys are known for is just like it's like really great execution and you know this is like the 1% % better everyday thing. And this might be hard for you to answer as somebody because it's just the way that you are. but I'm curious if you can sort of speak at all to what it's like to run a company with execution as like, you know, a core excellence. You know, like I'm thinking of Amazon, for example, as a company that had to do this when the margins are when the margins are thin, like there's no choice. There are other companies, by the way, where like the sort of, you know, the the zone of genius has to be something completely different. They don't need to be great at execution.

33:46 actually they can just have periodic brilliant insights that are just so unbelievably stepchanging that you know you can actually afford to be sloppy and the types of people who are going to have those insights might be types of people who are going to more likely be you know le less you know attuned to the details anyway but I'm just curious if you can speak at all to you know what your experience is like building a company with sort of these values. Yeah, I yeah, that is a hard question. I mean, I would say, you know, it it starts first and foremost with a a love and appreciation for how math and humanity come together because and what I mean by that is when I think about the Door Dash business, yes, you you're right. There are a lot of metrics, there are a lot of constraints, low margins.

34:39 and and and therefore you have to be very good at measuring a lot of things. That's the math part. That's the how do you you know take a multivariate problem and make the best set of you know trade-offs and and calculus. But underneath it though is the recognition that on every single order we do, we have at least three humans involved. at least, you know, we have at least a dasher, a merchant, and a consumer who all participate, you know, to make something productive happen inside that city. And you kind of have to like both. It's not good enough to just be very robotic and all we're going to look at is the numbers and if the numbers are good, we're good. And but and if it has a negative consequence on somebody, then so be it. That's not good enough in my book. My you know my book is you have to recognize that if you look at the merchants right when I think about my mom who worked inside of a restaurant this is life this is not a job this is not a 9to-5 or oh what are you going to do from this career to the next career no this is every single day my livelihood every single day it's my identity it's certainly my professional income but it's everything in in in the household you look at couriers you know we have tens of millions of couriers who've you know delivered with us and they we are like a stepping stone for most of them. The vast majority of them are you know doing only a few hours a week and that's because they're they are trying to strive towards you know becoming a doctor, a nurse, a realtor, a teacher etc etc etc and and and then obviously we talked about you know the benefits to consumers and and so you have to have that appreciation and love.

36:28 If you don't have that for either the the the people or the math, I think it's a very difficult game to sustain because it's just not choosing the right game for you. It it seems like it could be rare to have both of those in one person, but you've obviously got a company full of I presume people who you believe have both of those things. So what do you how do how do you figure out if somebody is not only one or the other but somehow both of those things where they appreciate the humanity and all these complexities while also just being you know a maniacal sort of stone cold you know operator when when they need to be too.

37:07 >> Yeah. Look, I I I think the test towards you know, someone's skills or their problem solving or their metric orientation is a lot more straightforward to assess than say someone's values I would say or and and and and and there a lot of it is actually hearing about the things that motivate them as well as the things that demotivate them. And it's okay by the way there's no judgment here. It's really around self- selection. Some people I think find it awesome, you know, the types of businesses that want to become restaurant tours, retailers, grocerers, some find it messy and not for them. That's okay. That's really okay. And and and and so again, a lot of this is about self- selection. It's about you about people who are going to do the right thing even if maybe the numbers bely you know that behavior people who have seen adversity people who believe in the fact that you know if we can be successful then all these awesome creators and passion projects inside cities will actually continue to be successful and actually you know want that to be successful. I think if people can self- select into that, that's really how you can tell. There isn't like this perfect test, but it's really around self- selecting into it.

38:34 >> That's cool. I'm curious like for you when you think about like growing your business, what are like the biggest levers? Is it is it like city expansion still to some extent? Is it like is it now about broadening out through more categories? Like is it M&A? Like what are the things when you're like, I want my business to grow by X amount next year. Here's how I'm gonna ladder my way there. Like at this stage, obviously a very mature business. What goes into it?

39:02 >> Well, well, I don't know if we're a very mature business. I mean I I mean relative we've we've we've certainly surpassed you know the the size of my apartment but but but I but I would say that you know we're still even our largest you know business our restaurants business in the US is only singledigit percentages of the of the restaurant category but but but to answer your question I think it's how can we either solve current customer problems better or how do we solve the next problem and what is that next problem and you know I think a lot of times you know back to the comment you were making earlier about some of advertising teams at some of these larger technology companies.

39:40 >> I I I think you you have to do two things. You have to keep building the core which you know for us has always been food and just constantly work on that problem of improving selection, quality, price and service. And then you also have to create the new which is actually a very different set of skills. It's a different management system. It's a different people sometimes certainly different incentive mechanisms it's it has a lot more inefficiency before you have efficiency and and that's where we're searching for new problems to solve and this is your point around you know well how much do you focus you know versus you know just doing the core it is both I think when you look at you know companies that can continue to grow they tend to do this they tend to keep solving the problems that they've solved for customers continually better and they also find new problems to solve.

40:38 >> How much do customers care about speed? Like if you could deliver everything in 5 minutes, would that would that dra like can you tell if that would drastically change demand or is that no longer a huge variable? >> No, I think I think people I mean this is like saying would you like something delivered slower? Right. People are always going to want something delivered faster. It's all and now is it is it x minutes for like depends on probably what the product is.

41:02 >> Yeah. >> But but but but the short answer I know is definitively is that customers always want something faster. >> Do you think drones will happen for this kind of delivery? >> Of course. I mean like drones drones, autonomous vehicles, they they will all happen, you know, but but we have to remember something. The delivery part is just one part of the time of a delivery. Yeah. >> Right. So, while it may be possible to fly in the air and skip a bunch of traffic, it's not possible to skip a busy kitchen, especially if you're understaffed. And so, and that's the predominant part, the majority time spent on a delivery is the preparation, whether it's inventory inside of a retail shop or, you know, cooking time inside of a busy kitchen.

41:43 >> Yeah. And the drone can't like go through the retail store and check out. >> Yeah. Yeah. Yeah. Yeah. So, but my perspective but but but again I think you're you're calling out a very good point which is when you think about you know products I I think when it comes to digital experiences a lot of attention is around pixels and every you know detail from step to step to step >> no different in the physical world but a lot of the steps are just in the physical world right >> and and and it's the coordination and the orchestration of the endto-end system you of which just the fulfillment is one part, right? Understanding exactly did you get the right inventory, understanding the exact prep times, understanding which vehicle to send, you know, whether it's autonomous vehicles or human drivers, understanding what is the right price point, understanding how do you solve, you know, substitutions, understanding how do you do refunds and credits, all of these things have to be orchestrated and I mean effectively hidden in terms of the complexity behind the surface to offer a very simple just get you exactly what you want, you know, to the customer. It's funny like as I'm thinking about it's like both you have both like the there's like the simplest kind of you know objective like a company of this scale could possibly have which is like get that item from that place to that home as quickly and cheaply as possible. Like that's really simple. But then like the complexity from A to B is ridiculous.

43:04 >> Yeah. >> Yeah. >> That's I mean I mean and that is the Door Dash problem set, right? It for better and for worse. I mean, there are, you know, I remember when we started Door Dash, I remember decomposing that there's almost like 20 mini systems, many little processes, if you will. If you imagine a checklist of just bring you a burrito, there's 20 little things that you kind of have to get right because if anything goes wrong, you know, you would actually need to fix it.

43:31 And those are and and and there's no way, by the way, that that that you would know about this unless you actually did the deliveries yourself, right? This is exactly why we still to this day, I mean, we've done it since day one, but to this day, we still have everyone in the company, you know, dash, which is our way of saying doing deliveries. We still all do deliveries because until you actually get into the physical world, until you find yourself stuck in the wrong elevator or the wrong lobby to try to get upstairs to deliver something, until you get to the wrong alleyway for parking, until you get to the wrong place because you realize actually food gets made sometimes in three different stations at inside of a restaurant. Until you actually experience those things, it's very difficult to op priori just intellectualize that and know about it.

44:19 >> Yeah. What are you most excited about for whatever's coming up with AI or or anything else in the next year or two? >> Well, I think right now it's a period of rapid change for everyone. And so what is very exciting is you know we started this conversation talking about some of the you know inefficiencies that perhaps companies are experiencing when it comes to token consumption >> token mixing >> token maxing and experimentation.

44:50 >> but there's a lot of fun in that too >> of course >> you know >> and there'll be a lot of creative like you know new discoveries that will come. >> Yeah. And so what I'm excited about is all the things that haven't yet happened. Actually, that's what I'm really excited about. Not not just in like products that that that will be great for customer outcomes and and and in increasing the surplus, but but actually ways of working. I am really excited because, you know, I think one of the things you always ask yourself as an entrepreneur is how do you continuously keep up the velocity and pace at a company similar to what you had when you started the company?

45:26 >> Yeah. And it's very hard as you know, you've done it yourself and and you see a lot of startups today and and and so I'm interested in answering both of those questions. It's like how are we going to take this period of change to yes build better products for customers, but also literally build better products for ourselves so that we can enjoy work to the max. Actually, now that you say this, I want to my last question is kind of on the personal because it it was, you know, I didn't do it for as long, but you know, nine years of it. And, you know, it it's difficult and you're, I guess, 13 years into this company and obviously it's, you know, was and is very intense, but has your own, you know, when you think about how do we keep it as intense as the early days?

46:06 Is that kind of the goal in your head or do you at some point you know you're now you know whatever many billion dollar public company do you at some point say I actually now need to find the marathon pace that I can do this for the rest of my life or is it just the same intensity as day one for you? >> Well it's it's less about like the intensity in terms of like you know how many hours are you you know sprinting or I I I think it's the feeling of agency and productivity that actually that you always yearn for. It's not I I don't think any startup founder at least I personally know is trying to optimize towards the number of hours worked per week or something like that. No, I think what when when when you know the reason why people go towards startups is and why they want to have a is because they want to have a lot of impact and and it's the feeling of the agency.

46:53 >> Yeah. And I think, you know, I want to make sure that everybody at the company you know continues to feel that agency to hopefully do more and more and more because you know what because because not only then will they do the be best work of their careers at Door Dash, but even if and after when they you know leave to go pursue whatever their next endeavor is in life per personally or professionally, they're going to have more confidence to do it.

47:19 >> That's awesome. Well, Tony, this was a pleasure. Thanks a bunch for it with me. >> Yeah, thanks Jack. >>

Summary

Tony Xu discusses the challenges and opportunities in leveraging AI for DoorDash, emphasizing the importance of solving real customer problems while managing the complexities of logistics in a physical world. He highlights the need for companies to balance efficiency with customer-centric outcomes, the role of AI in enhancing productivity, and the ongoing evolution of consumer behavior in the food delivery market.

- Companies are grappling with how to effectively utilize AI budgets to improve customer outcomes.
- The discovery phase of new technologies often leads to inefficiencies, but focusing on customer needs can guide successful implementation.
- DoorDash's AI initiatives have improved merchant onboarding and fraud detection, showcasing measurable benefits.
- Xu emphasizes the importance of understanding the complexities of logistics, which involve multiple human interactions and unpredictable variables.
- The company is exploring autonomous vehicles and LLMs to enhance delivery efficiency and customer experience.
- Consumer trends indicate a growing preference for convenience and speed in food delivery, with a shift towards restaurant spending.
- Xu believes in maintaining a balance between core business improvement and exploring new opportunities for growth.
- The company culture values both data-driven decision-making and a deep appreciation for the human elements involved in the delivery process.

Questions Answered

What is the significance of AI spending in companies like Uber?

The discussion highlights the challenges companies face in effectively utilizing AI technology to solve real-world problems and improve efficiency. The speaker emphasizes that technology should ultimately aim to make things cheaper and more effective, but the rapid consumption of resources, as seen with Uber's AI budget, raises questions about the tangible benefits gained.

Why is it essential for technology to engage with the physical world?

The conversation underscores the need for technology to extend beyond digital interactions and facilitate real-world actions, such as moving items or providing services. The speaker argues that software should ultimately serve physical needs, emphasizing the importance of tangible outcomes in technology's evolution.

How have consumer spending habits changed over the decades?

The speaker discusses the shift in consumer spending from groceries to restaurants over the past 75 years, highlighting the increasing preference for food prepared by others. This trend correlates with the rise of dual-income households, indicating a broader societal change in how people value their time and convenience.

What are the key challenges in balancing advertising effectiveness and consumer experience?

The speaker reflects on the complexities of the advertising business, noting the difficulty in creating ads that yield high returns for advertisers while also ensuring a positive experience for consumers. The inherent conflict between these objectives makes it challenging to maintain a balance that satisfies both parties.

What strategies are considered for business growth in a mature market?

The discussion focuses on identifying key levers for growth in a business that has already achieved significant scale. The speaker emphasizes the importance of solving existing customer problems and anticipating future needs, suggesting that innovation and adaptation are crucial for continued success.

© transcribe · For agents Built with care and craft by Gokul Rajaram