Uber Vice President, Marketplace Engineering
UC Irvine MS, Computer Science
Current Time 0:00
/
Duration Time -:-
Progress: NaN%

How did you get to where you are today? What is your story? What incidents and experiences shaped your career path?

Summarized By: Jeff Musk on Wed Jan 29 2020
just with a bit of an introduction on dhe to frame this, Uh, so I'm 35 years out of college. I spent my twenties thinking I wanted to be in academia. I started a PhD program on my thesis advisor. Took his research team to NASA Ames here in the Bay Area. Which is how I got to the Bay area. And I spent a number of years in my twenties thinking that I wanted to be in academia on Dai published in papers. You can look me up on Google Scholar, and, um, I did your machine learning back before it was cool. Ah, long, long time ago when it was much more of an academic discipline and was not something that was in demand in industry on for a variety of reasons decided that I wasn't interested in the academic rule on We can come back to that a little bit more, but I, uh, decide I was more interested in applications than I found it. Not as satisfying as I expected to to do a, you know, to present a paper in an academic conference, that kind of an abstract thing. Like I said, machine learning was a bit abstract and not applied back in. This is the late eighties, so it's a long, long time ago. So I decided that I was interested in, um, industry. And I was in my late twenties and knew that I had no experience, and I was actually threw somewhat randomness. I found it on a used net, but it was a job board. It was a very stealthy, anonymous, uh, request to hire people for a very stealthy company that wouldn't tell who they were online on. I jumped in the start up on, and it worked out really well for me. I often get asked whether to do start up or big companies when you're early in your career. For me in particular, that worked out really well because I I learned a lot very quickly. I was sort of willing to raise my hand to do anything, and I tried a little bit of everything. I was everything up full time engineer to Ah, I did pretty sales work. I did post sales work. I worked booths at a trade show. I wrote the first level let's document taped, and I went on site and some of our customers and kind of insulting workers were very small company. That was a really neat bootstrap for me because I really got to see a lot of different things. And you don't do that When you go to a large company, I often tell people in their twenties, you know, get a small company experiencing a large company experience. Did you get a very different things out of it? At a Google, you get incredible mentorship you get Ah, um, you see best practices you get, Ah, massive impact. You know what you're in product does because, you know, Google has a 1,000,000,000 users for a lot of products and things like that. But you get kind of put in a very particular neech. Whereas if you're a 10 person startup, you put on his many hat. You put on 20 hats in a week. Um, but a right I spent then and I'm ah, from my thirties on a series of small companies want the first of which was that startup I founded my own company in Web one point. Oh, in 1999 when no one really understood what the Internet was and was not for We got venture capital funding. And again, I learned a lot from that experience. Ah, and ah, different kinds of things, some of which is you never use again. I learned how to rent office space in which you don't use in those companies. But I learned a lot from starting a company from scratch, trying to figure out what the idea was building the froze part of Pro Prototype asking for funding, hiring the first people, et cetera. Did another company. And then I joined Google in 2004 just before the I Po. Um, and I spent 12 years ago Google doing a variety of rules, all in the ads team. The last six years I was in YouTube ads, which was a very rewarding experience. After 12 years, I decided Google was long enough, and I wanted one Maur, you know, really intense operational job. And so I took a job in uber three years ago and spent the last three years of Hubert, which I recently left, leading their pricing and matching teams. So there's a lot of different ways of answering, you know, sort of what instance experiences shape, your career path and we could spend longer on that. That was sort of 35 years in four minutes. Um, I often describe I didn't see like when I was 25. I certainly didn't imagine being a VP a Google when I was 50. I never probably looked more than two or five years in the future, and I don't know that's the right choice or not, but that is what I did. I think Maurin terms off. I made good decision. That's impossible to know if you made all the right decisions. But I made, I guess, reasonably good decisions that each mean pivot point. Um, and some of those boys, of course, Whenever you're looking for your next job, you know, when I switched from company extra company, why which happened maybe five times in my career. And I think those are the big decisions. That's when you leave one company. And when you decide to leave one company and then you decide which one to go to, and all you can do is really just get a cz many different offers in front of you as possible and make the best decision you can. You can never see the future perfectly. I'd love to say when I joined Google in 2004 that I knew a guy who was gonna become Google was already reasonably well known, actually, quite well known. It was the world's largest search engine, but it wasn't an operating system. It wasn't day browser. It wasn't display ads. It wasn't Google. Um, uh, Cloud. It wasn't a fallow, and it wasn't even Gmail yet. I think when I had my offer letter, it was a search company, um on. But I had an instinct that it was a good company and I jumped at it. Andi, I think things like that it's it's you gotta teach main sort of pivot point, make the best decision you can. Um, and I think the other big decision I talk a lot about when I'm talking to people earlier in your career is the decision. I made a certain point Thio, whether it be on the individual contributor track where the management track about 3 43 years into the startup idea that I mentioned earlier in the nineties was I was in my early thirties, relatively new into my commercial career. Ah, I was a good engineer a TTE. The risk of sounding here again, I, you know, was a good programmer. I was able to make a lot happen in a short period of time, but I had the opportunity to try a leadership role, and that was again. One of the things that I think was good about being in a start up is that, you know, bad opportunity availed itself because it was the start of that was growing, and I stressed about it immensely. I think I spent several months kind of try and decide whether to do it. But then finally someone said, Just go for it. You know, if if he doesn't work, it doesn't work on Di did try management. And I think that was the right career path for me the last 20 years. I think it Maur is suited to me. I don't think everybody should be a manager at a little, and I tell the people that all the time I have a lot of people come to be three years out of college saying I want to try management and a my answer is you too early in your career. You're too early in your life, be managing responsible for other people. That's there's exceptions to that. But I think that's broadly speaking true, and I think it's good to get experience as an engineer before you try to manage engineers. But I also think there's some people who don't want that responsibility and it doesn't match their personality well, I think it did match my personality well. And, you know, I spent the last 20 years of my career managing other people, and I really enjoyed that experience.

What are the responsibilities and decisions that you handle at work? Discuss weekly hours you spend in the office, for work travel, and working from home.

Based on experience at: Vice President, Marketplace Engineering, Uber
Summarized By: Jeff Musk on Wed Jan 29 2020
So as I said, I left this job four weeks ago, but we'll use the present tense. One of things I loved about and my role and head of marketplace engineering and uber was I led a 400 person engineering team that was cross functional, about 600 including data science and product management. Some other rules that didn't report to me and one of things I loved about getting to that level. And it's pretty different from a job where you're managing a first line manager of an eight person team, which is, you know, a lot closer for people earlier in your career. But I like the variety. Things were just different 20 times in a single week. There was no one thing I did a lot of, Um, you know, some of it is personnel. You know, when you run a four person team, there's always something. There's always someone who's unhappy or you're trying to keep make happy. There's always two people who can't see eye to eye on some personnel or technical issue. Um, there's always one person who you hire that maybe you shouldn't have. There's always someone who has a family reason that they're, um ah, not ableto put all their all into work in a given week on. And I, um you know, I only had maybe eight or nine or 10. I don't remember direct reports, but you know, the more difficult problems bubble up to you, no matter how much you try Thio, um, delegate things and trust your leadership. There's a lot of personnel things around running a 400 person team. It happens. I just was talking to my successor a couple days ago, where there's an annual performance review cycle where we decide who gets promoted or not, and who gets what size raises and who gets what. Size bonuses. And that's probably no managers. Favorite part of the job. It's a painful thing, but at the same time is incredibly important to kind of do those. So there's a lot of just personnel work, and anybody who thinks management is all about power and not responsibility should to think twice and or if you experience it, it's it's quite different. There's always, you know, some challenges in that area, but the areas I really liked and probably enjoyed got the most satisfaction out of were you know, I was responsible for a very large, functional team that could make an enormous difference for Uber's customers. I lead. As I said, it'd probably three or four main areas, the core trip flow of uber. So the system that ran the state machine for both the riders and drivers who were in the middle in process either driving around his drivers or riding his riders. I led the pricing team trying decide how much to charge the rider and how much to pay the driver for 20 million rides a day and matching team, which is assigning riders to drivers. It does a full scan of every city every five seconds roughly and does a big matching problem in all those places. There are a lot of technical decisions to be made, and leading a four person team is a VP, usually making a few technical decisions possible. You should be relying on others but having a good enough sense of what's going on in all those areas that you can have a sense of. When things were going well, when things were not going well, like if you run 10 teams just for the sake of argument. There's always gonna be 96 projects that are going great. Three projects that are going OK on one project that's floundering and your job is to understand which project is not being is not going well. Understand when you need help, not necessary. Dive in and make decisions, but, you know, support the team unblocked things in a place like uber, you know, there's a lot of policy issues. Obviously, we're constantly dealing with regulatory challenges that a lot of different company, a lot of different cities around the world London just change the rules in one way California change the rules in another way, and sometimes my value added, just going to a lawyer and saying, You've been thinking about this for three weeks. We need to get past that point because there was a recent situation. I can't go into detail where we have, like, three months before some regulatory rule went into place and the lawyers were taking two months to decide what to do, leaving the engineers like an hour and 1/2 to actually build it on. And I had to sort of help unblock those kinds of things. Um, so I would do a lot of project reviews, that a lot of one on ones and my job was to really it first proclamation know everything that's going on in my organization. So I knew where things where I could help on also and sometimes help men. Like I said, direct intervention. And then sometimes it's coaching people. Ah, lot of my product is a VP is the team, and my product is my five director, six senior staff engineers and 30 front line managers, or something like that and helping them all be successful because I can't do everything, but it's knowing enough to know how you can help other be successful. Um, it says it says that that discuss weekly hours. Uh, so first from philosophy perspective. Ah, I very, um, strong proponent of not counting how long people are at work. Some of the most successful people at both you and whoever I worked with on. I think this is different to start up. I think start up, you're forming a team. There's a very tight timelines. I think everybody being in the office for a lot of hours matters a lot more, but we're in a large company that's distributed my engineers and bang galore and injures in Seattle and injures in Toronto. There's people doing videoconferences from home. There's people working off hours, et cetera. I told people very early on I would not put a premium on how often you're in the office. I showed up every day, by all means, but I was probably in the office seven or eight hours a day. And then I did a lot of other time, you know, I would roll out of bed in the morning and check email to see that something had happened after I'd gone to bed or overnight. Um, I would check me out before I went to bed, so I was working often. But I, um

What qualities do you look for when hiring? How do you interview candidates?

Based on experience at: Vice President, Marketplace Engineering, Uber
Summarized By: Jeff Musk on Wed Jan 29 2020
So I interviewed leadership roles more than I did individual contributors because of the sort of role I had. But I think that the answer is similar. And I did interview, Um, quite a few sort of senior individual contributor candidates in my time at both Google and uber attitude matters. A tremendous amount it comes across and Google especially, and Newberg Jewelers areas. Well, you know, we got the best crop of candidates anywhere in the world. But we would notice a lot People who came in with an attitude, people who came in arrogant people, came in with, you know, if in a 45 minute interview, you already give indications you're not somebody I'm gonna want to work with, and that's when you're supposedly on your best behavior versus Went under, you know, very difficult situation at work. Ah, Google would try. It would generally spit those people out in the interview process, and that wasn't my intent. The rule that Google was if you didn't want this person next to you in your cubicle on your team, you couldn't recommend you can't say, Well, that's a great asset for Google. I don't want to work with with him or her. But, boy, I'd really love him working somewhere else. Person seems like kind of a jerk on. And I saw a lot of really, really talented people get rejected that way and go who was not perfect. Any interview process is in perfect. It's human. Um, but I uh um uh, so attitude really comes across, I think, humility. A growth mindset. Meaning just evidence that you still think you should learn. I think that our industry is so accelerated nowadays that everyone thinks their senior software engineer four years out of college and their staffs offer engineer five years out of college. And if they're not a CTO before they're 30 there a failure. There's a little bit of that feel in the sort of bay area hypergrowth mode that we've been in because we've been in a boom for 10 years straight. Anybody under 30 has never seen anything but a boom. Um, there is a ah, and I know I sound like a little bit of an old guy talking about millennials, but I think that it comes across sometimes that there's people five years out of school who think they've already learned everything there is to learn and they have, they just happen. Um, and Ideo I often tell people I'm still I was still trying. I'm still trying to learn now. Um and I was trying to learn six months ago when I was in my final job. There's lots of things I wish I was better at That comes across a lot. Uh, and I think it's what kind of this is all have the soft things. I'll come to the sort of hard skills in a moment, Uh, a desire to learn. I often counsel people, has a slightly different question too early in their career. And by that I mean sort of 1st 52 15 years. You should be optimizing for experience and learning and not for money. I told a story recently. I was speaking to some high school students of I was trying to close a candidate tojoin uber. I think a year ago on he was trying to side between us and lift, and whatever decision gets made is fine, but I'll make up a number. But like, roughly speaking, it felt to me like he was basically if we offered him $50,000 a year on uber offered. Lift offered him 50,000 and $1 per year. He got a lift. I mean, he was completely stacked. Ranking on compensation and money doesn't not matter. But early in your career, so much more. And I really almost wanted to withdraw that candidate. It was so clear that his like focus 123 and four with compensation. He didn't ask you what manager was gonna be. He didn't ask if the work was gonna be impactful. He didn't learn. Asked what the software stack would be, he didn't ask what kind of organization I was trying to run. He didn't ask about uber's future. It was like, Tell me my calm and like making more And I just was so turned off by that candidate. In fact, that candidate Gigio off the lift, as it turns out, because they offered him $500 a year Maur which I mean, it's not nothing, but in the great scheme of thing is nothing for somebody four years out of college. What matters? A lot more of you building your career for the future s o. I look a lot for those kinds of things of someone who wants to contribute once contribute meaningfully wants to learn on that kind of thing. I shall say one more soft, Gil, We I see a lot in the market nowadays of people's hopping jobs every 18 months. And the fact is, you can get awarded by that usually gonna raise when you switch jobs. But if you're 26 you've already been a four companies, it really feels to me like you've never had long enough to really become an expert in anything. And you might have made your compensation more about your resume. Looks a lot worse. I know exactly what the right sweet spot is, but by and large, and everybody once will go to a company that's a mistake and leave after six or 12 months, and that's okay. But if, like I interviewed, a guy wants you literally spent, like two years in today, it lift are now two years a day that a day of interest and then two years in a day at slack and then two years in a day at Airbnb and now is applying an uber and I'm just like, you know, I don't want to hire you here to be here for two years and a day, Um, in terms of hard skills. So again, I'm less the person who's asking you to code on the board, but a lot of how we look and in interviews and artificial process, and you get a lot of people argue about whether saw House offer interview should and shouldn't be done. But, you know, software's a team sport. And if someone can express their thinking, process can have an interactive, you know, when given a problem, ask clarifying question. Think in front of a whiteboard. Um uh, explain the assumptions they made and said I could do this, but I think you know, I got it right here by hand longer. I beat me do that, etcetera. That sort of thing matters almost that much is whether you can write in tactically perfect c++ goat or go or job or whatever that stuff like ideas, convicts like everybody makes angle bracket failures here and there. Somebody's gonna forget the order of arguments on S t r l e n Under the pressure of an interview, etcetera, Um, those things matter a little less to me, especially a world where so much is available online. It's it's your thinking process and your evidence that you can communicate because you'll find in most riel job situations. The percentage of its work that you do that is interacting with other primates as opposed to interacting with a keyboard is, I think, more than a lot of people understand from their academic experience, and, uh, so

How do you inspire and motivate your team-members? How do you foster creative thinking? How are ideas shared and implemented?

Based on experience at: Vice President, Marketplace Engineering, Uber
Summarized By: Jeff Musk on Wed Jan 29 2020
so I'm a big believer. So both of last. Actually, all my jobs the last 15 years have been very much in a business context, like fairly close to the customer. Not like leading the storage and infrastructure or networking team, which are incredibly important roles. But they're the customers are a lot more internal, and the metrics kind of feel quite different. Um, running teams liken, adds team or a matching team. I had front and engineers were building part of the uber Ryder and driver app, or part of the YouTube watch player on and things that were pretty close service is that were really pretty close to that. And I was a big believer in giving people as much business context is possible. I got finance mad at me all the time at YouTube because I would put up slides with all the numbers of how YouTube was doing. I finally would like hide the Y axis, but at least be able to show people like what's doing well and what's not doing well and I think especially near the top of the stack. I think the Maur engineers understand the context of what's doing well and what isn't on DDE. That doesn't just mean revenue. That means user satisfaction things, measures of driver happiness. To the degree that we could say at the end of 2017. There's a lot of work and uber that our drivers were X percent more content with their job driving uber, which is clearly tied to turn and people dropping out of the network. Um, I think people have a sign, meaning engineers want to write Good co. They want pretty, uh, elegant code. They want good unit test coverage. They want to know that they're, um ah, you know, proud of what they did. But I think that the thing that motivates a lot of engineers and the sort of people are successful in a job like YouTube ads, ruber market place or the people who are really motivated by the difference we can make with the business on with the experience of riders, drivers, YouTube creators, YouTube viewers, et cetera. And so I put a lot of work I used to dio in both of my last two jobs. I called them brown bag lunches, which the dated term, but like very informal lunch is where I just sit in a room with 2040 60 people, depending on the roll on and talk about what was going on. Talk about, you know, it is a VP. I'm in meetings with CEO or other, you know, high, highly. Ah, title people. And I would be able to say, OK, this is what's on. Our CEO is mine. This is what's she's worried about right now, like we've been doing this. But we don't really know how to solve that and try to engage my team, not just in. Please implement this stuff that we've passed down on high in stone tablets. But I want your ideas on what we should be doing now. You're not only your ideas are good, but it's not. You know, we do have a management process with privatisation process. We look a little setting process, all these kinds of things, but trying to get everyone on the team engaged in, um in solving the problems of business on. So ah, lot of what I care about was transparency and not hiding when things were going badly, because there's always something going badly, and in fact, I think that's what actually motivates a lot of engineers of like these eight things were going well, but we're having a real problem with this. And I want the smartest people on my team thinking about that problem and coming to me with a Kevin. We should be working on this as opposed to something else, um, and iconic ties into the creative thinking It it's it's people. But there is a role, and wherever you go, there'll be some sort of process of a product manager. Program managers called different things at different companies passing a requirements document to you, and that's a valuable role. I'm not trying to take anything away from product manners. I think they should be your product manager spent a lot of time with lawyers with policy, people with public relations people that engineers don't and they end customers. I'm sorry, actually, especially with customers, and I like getting engineers in front of customers directly. I think it's super motivating to have a human being in front of you who is a driver and saying this is the problem I've got, um I wish you could solve this for me, but sometimes it's I want more money, and that alone isn't sort of super insightful, but you've got a lot of other nuance of predictability of earnings versus, you know, people want their earnings to be. This is opposed to that when they're feeding their families strangely enough, and knowing that intellectually is one thing but having a human being in front of you or watching a recording of a human being because he might be around the world. I'm telling you, that matters a lot. But, um, to the degree engineers have the context of, like where these requirements are coming from, no, the priorities, et cetera. I think that leads to a lot of out of the box thinking. But I also we tried toe have, um uh, they're they're they're CEO. You might have heard of your Facebook has what they call Agathon. Google tended to call them innovation weeks where you and I would tend to do them like just before the holidays, because you can't launch any software the week before Christmas anyway, Are you shouldn't always says we Google in uber work on your kind of, ah breathing space of generally you don't have any December 20th deadline that ring here we would have a week where we would structure it That, um, we kind of had ah machine around this Where, um, anybody on a team? One of my 50 person teams that uber did one of these very successfully every year. You know, anybody could submit an idea. We want to spend a week trying to build this. We think we could build this. That would help our drivers be happier or riders be happier from and then we'd have a process where teams will kind of vote and you can work on anything that was sort of your week. But your goal was to kind of win the hackathon, and and so teams will coalesce. And rather than having 40 random ideas, that would be a kind of process of coalescing around five or seven ideas that seemed really promising to a bit of, ah, voting slash crowdsourcing kind of process. And then, you know, four person teams we call less because one person in a week on a big software stack, Huber's usually can't do a ton. But if you get five people really focused for a week on, maybe it's throw Waco. But just like I'm gonna build a prototype of this on. And then, you know, we have a presentation at the end of the week on a Friday where some leaders would and I would sometimes be a part of this process. Would C five different demos Ah, a different user experience, a different matching process. We've run it through a simulator. It solves this problem, you know, causes this to metric to get better this metric, get worse, and then we would sort of our commitment as leaders is that wasn't just a one week vacation that we would really make an effort to get some of those ideas. Be part of our road map moving forward. And some of you might have heard of 20% time a googol. It's a similar idea of giving engineer sometime that's their own. That's not passed down from above to do creative things, and then you've got as a leadership team and Google much more so 10 years ago. But you know, people scholar Google Gmail on some other products came from one person building a compelling prototype on, And so I think giving that kind of support to enter years having bottom up ideas matters a lot

How do you set targets for your team-members? How do you measure their progress? How do you incentivize them to meet their targets?

Based on experience at: Vice President, Marketplace Engineering, Uber
Summarized By: Jeff Musk on Wed Jan 29 2020
level. I don't incentivize. I don't set targets for individuals. I set targets for team. My team was broken. Let's just say for the sake of argument, maybe 15 sub teams, um, in from 10 to 60 people and I could digest at my level each of those people's top goals. Um and ah, uber tended to be doing semiannual planning. Googled a quarterly landing on there's different like level. There's large like there's a big annual planning process, which tends to take sort of October November, and that's both Google and Uber works somewhat similarly this way. Where, you know the CFO is trying to side with the budget is for the next year because the CFO of the CEO are trying to say, Okay, we're gonna grow engineering X percent next year. And then there's a process where all the components of engineering are. Frankly, it's a zero sum game. At some level, you're only gonna hire 100 engineers, and if you put 80 on one team, there's only 20 left for another team or whatever, and that's true it any size of company, um on, and that's a big process, and that's fairly heavy weight, and I we were just doing that. An uber in my last couple months, they're letting all the way up to the CEO, hears the things we want to prioritize and all the way to the top. They'll be strategic input on. We want to put more work into this in 2020 because and I know what it was. 2020. Priorities are. I can't really describe them out loud, but, um, we're gonna focus on this this year and therefore this team's gonna get a lot were head count because it's really important that the pick out meeting slots to fill engineers on a sort of micro level at my level, I'm meeting with each my 15 teams and you probably either have heard over can look up. Both Google and uber worked on Okay, ours objective results, which is, and I think originated at Intel in the seventies or eighties, but it's basically it's it's a way of structuring your goals, and I would review the top level bowl And my, um, uh, accountability mechanism was at that level like I would meet with a tomb and I would say this team is really knocking out of the park there in trying to prove this metric by 3% this year, and they made 7% on. And I would be more inclined that if I don't know which of the 60 people on the matching team, we're part of that success. But I know the matching team writ large with race except their leader will be I, uh, uh, rewarded appropriately for his his or her team success. And then within that team, it's up to him to kind of having an idea who within his teams were really the rock stars that helped, uh, over over achieve on their their goals and, you know, sort of There's different kinds of teams, like the matching team, which I've talked about a couple times, which is again measuring matches, riders and drivers. That team happened tohave really clear metrics that we all agreed on of what considered success. If you know, if we were keeping our driver's busy 52 hours of 52 minutes every hour instead of 48 minutes every hour at our scale, that saved the company $100 million or something like that. Well, you don't quote that, but, um very, very large sums of money. Those kinds of teams are really easy to measure. Their job is to make our drivers more efficient, and our riders have lower e. T a s that there's still nuance within that of If you improve one by a little bit and lose another one on the little bit, you know, how do you make those trade offs? But by and large, it was pretty easy to measure success. A lot of other teams are a lot harder to measure, but generally speaking, we would do semiannual planning. We would set a set of goals that had to be very, very objectively measurable to the degree possible. So that, like a failure, is if you set a goal on January 1st on July 1st, you're arguing about whether the team achieved that goal or not. You wanted to be pretty clear. You were supposed to do 5%. You did four and 1/2. It doesn't mean you get fired, but it means it was You're less successful than you could have been. And if you did 7% then you're doing great. Um, on and incentivizing people to meet their targets. I mean, obviously you know there are financial rewards. You know there's promotion, there's larger bonuses. There's more equity awards in a company like uber, but a lot of it was just making it clear to them, And this goes back to the business. Context I talked about a couple questions before is I want them to understand why, if they improve this metric 5% it matters. And if they know that improving that metric 5% makes drivers lives better and makes them take home more money for their Children, you know I don't really need to do much. So my job is to have them understand why the goals are the right ones. And then, um, it's the rest was easy now. I don't always succeed in that. Some teams have clear goals and others, but that's that's an ideal world for me.

What major challenges do you face in your job and how do you handle them? Can you discuss a few accomplishments?

Based on experience at: Vice President, Marketplace Engineering, Uber
Summarized By: Jeff Musk on Wed Jan 29 2020
super went through a challenging year in 2017 which I think is pretty broadly known. We changed out CEOs, got a lot of publicity that was less attractive than it might have been. Um, my job in a world like that and that was on Google went through various times in my 12 year career, where it was either broadly beloved or slightly less beloved, based on whatever the press cycle was that the weak Facebook's been through up and down cycles. Um, some of my goal and the negative press cycle that uber went through started literally my first week. It was perfect timing. Um, on my job was really holding the team together and keeping the motivated and keeping them, uh, focused on the task at hand, focused on what they could solve and not the things that they couldn't solve on some of what I was I was able to do That was exactly, I just said was telling some stories of, Ah, Facebook's up and down because I knew a lot of people from their some of Google's ups and downs, you choose ups and downs most you probably already know. YouTube's owned by Google um, and give them sort of longview parts of just being a steady, you know, your blockers leaving lost his focus, your bosses focusing on the task at hand on. I was very proud that in that year we had attrition. Um, everybody always does. I mean, you know, the best company in the world, you know, x percent of people need Google the year after my p o. You know, back in 2004 but I was able to keep the team pretty stable. I didn't lose a lot of people that I really wanted to keep. Um and actually, in 2017 we actually achieved quite a bit. You know, it's never as much as you'd like, but I felt really good about that on so a lot of my job there was really team with keeping people. And then there was, frankly, a time in 2017 where the reputational challenges were such that recruiting was not easy. Um, I was able to hire some people, but a CZ things came out and a lot of ways that they will tell you teams just just be patient. You know, it's the press, gets bored of beating up on one company and you know, it was uber and 2017 it was. We work in 2019 united who it's gonna be in 2020 yet on a very good friend of mine was a senior leader. It we work, and we would commiserated over cocktails about just how much fun it is being that company for a period of time. Um and then in 2018 you know, we were out of that press cycle and uber became appealing again and are more appealing. And it became probably know we were working up to a typo on and a lot of my focus was on team building just really bringing an excellent talent. I mean, a lot of your product, as I think I said earlier before in person team is your team and I put an enormous amount of work into recruiting. I was doing Oh, God, I don't know five or eight interviews a week for an entire year or something like that. I did the most interviews of any uber employees in California. I was told in 2018 um, on and really having enemies being relentless and that man like I'm calling. Somebody said, I'm not interested in I say, Let's just get here for lunch on a Saturday You know, even if you're not interested, let's just chat. And one of those people ended up being one of my best hires. I one of my best hires, I literally found because he was another parent at my daughter's back to school night. I was recruiting at my daughter's back to school night when? One year, Um and so it's just relentless recruiting the talent wars in the Bay Area and sort of, I think, the tech industry at large. Getting the best people just requires an incredible amount of work. Um, I think that a lot and then another side, and then the other accomplishment I think I talk a lot about is just alignment. I think that when I got there, we were not nearly as good as I talked a lot about this and previous question about measuring success, and I think we got to the point. I don't by no means get all the credit for this, and I had a lot of partnership from people improbably event data science, where everybody agreed. What success Waas I think of time we left. And so my last meetings I went to where I was sort of no longer in the job of watching my successors the quality of conversation around. We all agree we achieved 3% in this metric in the second half of the year. We've done well from the business and, like, only focus then is just a lot. How do we get there? That was super gratifying to me.

What are the job titles of people you routinely work with inside and outside of your organization? What approaches do you find to be effective in working with them?

Based on experience at: Vice President, Marketplace Engineering, Uber
Summarized By: Jeff Musk on Wed Jan 29 2020
It's the same between Google and uber, and I'll talk about something that's different. Um, in most rolls Ah, one of your as a engineering leader and really his individual individual contributor as well. Um, you one of the most important roles with your product manager, um, most companies that I know of, product management is a parallel ladder. So there's, you know, a bunch of product managers who, you know, report up in one direction. This bunch of engineers report up in another direction. Sorry, my end of the phone, please. Um on and they need to work together on it is not always an easy relationship. There are always tricky questions of whose decision this is. This is a technical decision is it's a product decision because almost nothing is purely a technical ization of purely a product decision. Every products and for those who are mount familiar with this because they're in their early in college there L, a superior product manager, is typically the overly simple thing is, what do you do and the engineers get? How do you do it? But of course, there's tons of overlap in that, because there's engineering decisions that might be. This is a beautiful solution, but it's gonna take two years on, and there's product requirements that might require just horrible, horrible hacks and ruining of the architecture. And so there's very few decisions that don't kind of cross those boundaries on DSO. In every level, whether you're a leader, a manager, an individual contributor, a director of E P. You're gonna have a PM partner and they're sort of parallel universes. And I think building that relationship and having it be as much of a trust relationship where you, um, are trying to not just say no, I mean the sort of failure mode. Is engineer saying, Well, we only go go away on I've seen that level of dysfunction, even a Google. But if you can come back and say, you can have this But here's here's the downsides. These five things were gonna break. We're gonna need 5000 more servers, which you might not be thinking about. Our database scheme is gonna go to hell. You know any of those kinds of things. Two way communication of the implications of product decisions, I think is incredibly important on DSO, I think, and I just was out for dinner a couple nights ago with some of my best friends from YouTube ads, and they're almost all product managers. I think I had a I really worked hard on that, being a productive relationship because I think when product and engineering are working really well together, I think that's that's when teams really saying My Yuji Brad's team and sort of my first few years there was one of the best experience in my career there, Um, so I think those sideways relationships really matter a lot. The product one is probably one of the most important ones, and similarly, and it varies from company to company. But the customer relationship and your customers could be different in different places. I've talked a lot about getting right in front of you can't Hoover has four million drivers. You can talk to all of them, um, YouTube as a 1,000,000 users, but figuring out how to have good relationships with sort of representatives of those organizations. It was sales people at Google, um, it waas what we call the operations team at YouTube and really having engineering B is collaborative as possible so that there not just the people hidden in a corner who are hacking out code that no one understands, but they're a part of the business. Um, so I did it like that. It's funny. I just had breakfast this morning with someone who is a lawyer from uber. You know, large companies are really complicated, and all those rules are important. And I think there's I May 1 of most passionate advocates of engineering having a seat at the table at every company because they're the people who are building the software product, Uh, but without every rule, And that's everything from lawyers to policy, to product managers to marketing people saying, You know, you could build that, but I can't say I can't position it. It's gonna look bad. It's gonna you know, it's gonna blow up. Uh, you know, the New York Times gonna write an article about who goobers or jerks if you guys do this so we really need to do a different product. I think the degree you can get all those different voices because they're all important working together on, and I'm not saying I'm always perfect or it's always been perfect. All the teams I've been in, but that's what you should be aspiring to

What was the hiring process like for your job? What were the roles of people who interviewed you? What questions were asked and how did you answer them?

Based on experience at: Vice President, Marketplace Engineering, Uber
Summarized By: Jeff Musk on Wed Jan 29 2020
both Google and uber and mid stock. Coincidental. I mean, a lot of tech companies work pretty similarly working away that you'll have a panel of four or five people. There will be rules about who those people are. It can't be. All managers can't be all individual contributors. Ideally, it reflects a diverse slate to the degree possible. Um, you know, you put five white guys in front of Ah, candidate it. It says something about the company, which is not what you want to say. Um, and also diversity of talents. And typically there's modules s o i e can't remember off the top my head, But, like, for the sake of argument, they'll be one soft skills. Um, interview one on architecture to one coding and something else. Um, on. And that way you're trying to ensure that you don't have five people ask the same question, like if you go to a company and this this I mean, I'm probably with part of this 25 years ago. You know, if you don't really organize your interview process, you have five people ask the same question or basically the same skills on you. Don't really get the full signal, and you want as much possible have you know there's different cancers are excellent, different things. It might be great architect to their codes, not perfect or vice versa. Might be people who their ability to communicate A design is astonishing, but they can't write code at all. And really two degree possible. You want to get a full up full picture of the candidate you've got in front of U. S O. At both Google maneuver, there will be modules, and then there would be approved questions for those modules. Um, and you have to rotate through them because somebody or post on medium. I just interviewed a Google yesterday. Here's the three questions I got and people search for that Et cetera, et cetera. Um, abide I, um uh I would tend to be a person, And then that was for individual contributor rolls. And the modules were different for leadership roles, you know, because there is a lot more on leadership. Um, how you know how you operate operationally run an organization and things like that. Um so and then the and then those five people will all right, feedback and get in a room and sort of and actually the way Google did. Our uber did it. Waas um, you didn't want the first person talking. I loved him. And then that kind of swings everybody else. You do this thing of, like, orders. Where's my hand? You know? Here we go. 1234 You know, up, down. You know, whatever and sort of get a quick sense of you know, if if there's five thumbs up from five different people events, five different questions and it's a pretty short meeting. If there's five thumbs down that it's a pretty short meeting, it's when it's 3 to 2, which is actually what happens most of the time that it's a long meeting. Oh, you saw that. That's interesting. Okay, this candidates really seems the rock star at this, but really serious gaps in that what we do. And that's what you're tryingto figure out until be like a variety of different questions. For an individual contributor, all almost always coding at the board were designed at the board. And as I think I said before, what both of my last two employers would look for a lot is, um, call ability to think out loud. By all means, like you sit and think, You know, for 15 seconds after somebody pushes a present. A problem to you. But I think you know, interactive. Like when you ask this. What's the assumption? Did you mean that? Did you mean that? Okay, I'm going to think about this. I'm heading in this direction. Give the person opportunity, like, clarify something. Two minutes in communications. A lot of this perfectly syntactic code from someone who looks like they'll never open their mouth for three years if you hire them may or may not be exactly what you're looking for. As I said, Software's a team sport.

How did the program prepare you for your career? Think about faculty, resources, alumni, exposure & networking. What were the best parts?

Based on experience at: MS, Computer Science, UC Irvine
Summarized By: Jeff Musk on Wed Jan 29 2020
It's interesting. I get asked forms of this and it's been long enough that I'm not sure my answers are perfect anymore. Um, I think that it prepared me well for the early parts of my career, and that's what I said is the growth mindset. The field changes a lot. No one was talking about cloud computing when I started. No one was talking about big data. What I started, people weren't using Apache Spark when I started. I'm a big believer in the building blocks and I thought I got pretty good building blocks of, you know, atomic matters. Hardware matters, networking matter, data, structures matter algorithms. Mariner on. And I happen to be a specialist machine learning when I was working my masters in my early PhD work. So I got those basics, and I think those things matter. And then you're gonna have to like the sound of the programming language gonna change super fast in the next 10 years. But, you know, if you're all 20 now, the world's gonna look really different. Window 28. And so a lot of what I think you're looking for is, um, enough building blocks to get your foot in the door for that first job or two. And then it's gonna be on you to keep learning because things gonna change fast. There's a new framework. You know, You I frameworks change every year and 1/2 assed. Far as I can tell, I can't keep up with what is the flavor of the month, etcetera. Um, you know, looking back as I was talking to, I'm mentoring some Stanford graduate students right now. The degree of information you all have because the world has changed with linked in with Googling with etcetera with stack exchange, all those kinds of things is nothing short of astonishing. My network exposure, et cetera, in when I came out of school was like 1% of what any of you have. I certainly didn't get videos from deep ease of top tier companies to look at. When I was 20. I I feel like I was so much more in the dark. I had no idea how to find Top Gear Cos how would learn about companies, et cetera. It was incredibly more difficult in a world pre Google pre linked in pre Facebook pre did. That's that's, uh, um I think that and I applaud this'll program here of sort of giving exposure of people who are early in their career that folks have been doing it for a long time. I think those sort of things matter a lot.

Do you have any parting advice for students and professionals starting out in your field? What three mistakes they should avoid? What three things would help them the most?

Based on experience at: Vice President, Marketplace Engineering, Uber
Summarized By: Jeff Musk on Wed Jan 29 2020
trying to touch on some of these along the way. I should have these at AM because this is a common form of question. Um, a lot of it is optimized for growth and learning. Uh, you may think you know everything. That 20 or 22 I don't know what I'm talking, D'oh. But you don't on. That's just the way the world is being an industry. There's just a tremendous amount to learn. And I, um, to the degree possible, what matters the most in your 1st 5 or 10 years is like the experience you get and the people you meet. Um, I was told once I I'm horrified to say that I was 39. I was offered a job at Google. It didn't have a huge pay cut because it was pretty. I pO and I took a big title cut because uber was I mean, Google was very ah, um against title inflation. So everybody had, like, a title lower. I almost didn't go to when I was 39 which would have been a painful, painful mistake to make on a very wise mentor of mine who I'm still in touch with, said go to the right company, don't go to the right job and it was the right company. Opportunities will come now. That doesn't always work. Sometimes you go to what you think is the right company, and it's not but Google. I mean, I had a good career before I was the Google. My career with Google was a career changing job for me, and that's because I swallowed my pride and swallow my ego and bet on myself and said, I'm gonna go prove myself the jerks that Google and gave me tight election son of title. Why? And started me at this salary, which are worth way more than and I did. I proved myself. I busted my butt for my first, actually on my entire time of Google, but especially my first few years and was able to make my career on just learned a tremendous amount. Um, so that's one answer is outright for experience. I think what I did say, I think, try a small company and a big company early in your career again, not every year, but, um, you know, do three years of each because some people love what big companies air about, You know, you have 1000 personal hands and you're part of a big thing. And you know your household name, maybe, or you great things happen out of your company. And some people hate because you're, you know, one tiny part of a huge beast of a company. Um, you know, Google's got 100,000 engineers now or something just insane, like it's hard to make it in back, especially early in your career. Um, and some people love all the excitement of a startup where everybody knows everybody's name and you are all rowing on this. You know, you're on the same team and your first customer is a big deal, and all that could be awesome. And some people hate it because it's chaos. Almost always. It's just nothing has found it yet. You don't have a job ladders. You don't have a job descriptions. No one knows what they're doing. You're trying to figure out what's going on. Some people find that extremely uncomfortable. Some people thrive on it, and I think trying sort of both is really a good thing. Um, already touched on dough, overly focused on compensation. Money matters, but what matters more is the career you're building early on. Um, and I spoke to a bunch of high school students. I think I said two weeks ago it It is funny, but I think back more and more on my career and what both made my career successful and I've seen some other careers go sideways is the degree to which I don't want to see people like you, because that makes it a popularity contest. But the degree to which you are viewed as a good co worker, it's astonishing to me how much it matters, and that's going in my own experience. I had a job from 4003 as a not very successful company that was really struggling to survive. But I had one guy on my team who really thought highly of me, and he got a doll. But this place called Google, and he referred me into this place called Google because he thought highly of me and he had credibility there, and if he hadn't referred me in, they were really reluctant to hire experienced managers from outside the company. At that time. I would never gotten that job, Um, so the degree thesis may sound stressful, the people, but I really say often you're auditioning for your career forever. Um, I mean, tech is a massive, massive field, but especially as you get further and further, no matter whether you settle in Salt Lake City or San Diego or Silicon Valley, the degree to which your reputation follows you is unbelievable, and it still blows me away. How often, you know? I mean, it happens at my level, but it happens at every level. People from good companies know who the other good people are from their last company, and that's who they recruit in. So you're and and you're almost certainly wherever you go at age 21 you're not gonna be there for more than three or five years unless you're extremely lucky and you'll be looking for your next job. And the last 20 people who you worked with are gonna be the people who help you get your next jaw. And so don't be constantly networking, networking and sucking up. But you know, people who are lazy people who are jerks, people who cut corners, people who are, uh, certainly unethical. I had a case recently where someone lied about his resume. And I'm gonna go out of my way to make sure I've been asked couple times Why didn't you hire that guy like because he lied? You know, it's like That's a thing. And ah, and I don't like to punish people. But I also think that should matter. I caught him red handed, lying about his previous experience. Um, all those things, and it's sort of like the things you learn in kindergarten. But it's very, very riel of the way Silicon Valley works is the best. People are constantly looking for the best colleagues and the best company and to the re or one of their best colleagues. They will pull you with them. And, ah, just everyone is always watching you and your peers. You may think like my manager thinks I'm great. Doesn't matter to my co workers. Think of you that there's nothing could be further from the truth. Your co workers are gonna be your network, your roll index and, frankly, your references whether you want them to be or not. When you're applying for your second job, 24 you're gonna have three references. Who say you walk on water because we can always find three references who say we're great. One of them might be my roommate and another one's my brother, but, um, the good people, good recruiters, good hiring manager will go. Oh, you know, Kevin's coming from Company X. I know somebody company acts. I'm gonna just reach out and they'll text somebody and go, What was Kevin really like? And somebody'll go. Kevin was lazy with jerk. You sucked up to his boss, etcetera, etcetera. And, um, that's not followed you. So it really, really matters to do your job, right? And I know that sounds apple pie and obnoxious, but I have seen it happen so much as I look back on 30 years of watching this industry.