Sebastian Thrun on the Universal Law of Innovation: Build It, Break It, Improve It

first and foremost I'm I'm really humbled to be here today and it's such a distinguished audience I know for sure that a good number of you deserve this award more than I do and please be don't be mad at me you're amazing inventors and the talks have seen it just blow away interesting and great but since I'm standing on the stage I want to tell you a few words about what excites me what drives me and why I picked the port is that I choose and I want to start with my genuine passion that makes me get up in the morning is to try to solve big societal problems and I share that with almost all of you of course but the protests that I pick 10 reports I'm really emotional about and I really fell in love with just like a small schoolboy let me start at school this is my childhood it's not me watching those movies and I fell in love with cars at a very young age and with the idea of cars driving themselves and it was amplified when I was 18 I lost my best friend in a car accident in a split second bad decision for his friend who killed both of them and then I didn't do anything on cars for a while until 2004 when darpur the US government launched something called the DARPA Grand Challenge the dark man challenge was a robotic race sponsored by the US government who is so purpose it was to invent a car that could drive itself through the desert and they promised 1 million bucks to the team they could build such a car in the first year this competition the results were unbelievably shockingly awful the best team will drive only 7 miles and the average team out of over on a team would not even do a single mile in an open empty desert that's where my juices started flowing in so I can do better than that and I started joining hundreds of teams in the world to try to invent something entirely new for the self-driving car I pulled together an unlikely group of candidates to build my system called Stanford students rather than paying them we gave them course credit and in the next two months they would work really hard to build the first iteration in our first iteration was within such a short amount of time it was a miserable failure again he is our very first Drive and we got it stuck in mud and he'll shovel it out and a few weeks later these pictures were taken sadly by the New York Times who decided to print them but what characterized us and all the other teams was a willingness for crazy rapid iteration build it break it they did break it so Carnegie Mellon had a team that fantastic job and from UC Berkeley overseas from Stanford we got this video when we went out to build a car they could drive itself it looked utterly impossible but less than a year later five teams had succeeded building a car that could drive hundreds 40 miles here's ours and if you look very carefully in the very beginning in almost crashes moment when it swerved against a hockey vault can barely see it the the race itself was only these races that I think now are really a little bit historical even where the cars were not on their own followed by a government operated control vehicle and we could just see in a control room what's happening and get some footage from helicopters 130 miles seven hours these cars drove which in terrain that robots had never navigated before and just below seven hours our vehicle Stanley arrived from this growing race as the first ever completed our van challenge netting Stanford at the time two million bucks of which of course Matt installed I'm chair and propelled the robot itself into the and Space Museum in fact exhibition just opening right now on time and space in the and Space Museum and the mall you can see the car there was my very first taste of self-driving cars still a far cry away from what we do today and since this is 99 you are gonna emphasize a little bit the process that we did we had two things going for ourselves number one was a completely impossible problem but well-defined and that really helped and second was an eagerness to just build fail improve build fail improve and the truth of life is if you just build you you fail you learn you improve you build again and do this long enough eventually you succeed it's true for everybody it's a universal law I think so that's the law you have to keep in mind I think as you go forward every failure is a great opportunity to learn something new and there's nothing better than learning in life so when you fail you should be celebrating most of us don't now I got the chance to collect a few days later after I helped Google with Street View and Google Maps to gather the the most creative people in this field to build the next generation of self-driving car and very importantly before we hired a single person we spend hours on end ask yourself what is a great milestone and the reason why we did this is we didn't want to bring people in and say you want to solve the world or something again abstract you want something very concrete and we found something concrete Larry and Sergey and Eric and the concrete thing is we scoped out a thousand miles on California highways and streets and downtown's regular streets which said you're just gonna make the car drive everywhere and all these thousand miles and it looked like a complete impossible thing to do driving a soft driving car in traffic with other people around small children that's crazy so it was a great thing to try and we made prediction long it would take to do this that lasted between under five years and 20 years depending whom you ask and it took less than a year to get to the point for our first cars what kind of secretly drive for Palo Alto here's a video of a car driving to Colorado no one had a clue what these cars were some fella was a new wind generation energy system other stop it looking for extraterrestrial life the cars didn't quite drive flawlessly we had human drivers inside and about 10 months later we got to the point where we could just drive everywhere and the next video is just a set of snapshots taken from the car as it drives to San Francisco Monterey Los Angeles what have you this is a kind of complexity of situations the car was able to handle even two years ago see your Java coming along cars taking precaution highways as it was fast-forward tollbooths this is Monterey downton areas lots of pedestrians we've really gotten to the point it would be impossible and the methodology was the same and we had a very clear and press call and we reiterated as fast as we could towards this goal so if I look back a little bit on this project I let me show you the last one this is a an example I would like to have sound they'll be actually able to take an extra passenger and make this person the drive of a self-driving car are you driving so we're here at the stop sign cars using the radars and laser to check and make sure there's nothing coming don't die anybody up for a taco yeah yeah what do you want what do you wanna do today Steve I'm I'm all for talking all right well go get a taco to drive we're turning into the park go it creep along here does anybody have any money I've got money now I've got a little wallet there here and roll down your window in order burrito yeah okay very well how are you today this is some of the best driving I've ever done 95 percent of my vision is is gone I'm well past legally blind you lose your timing in life everything takes you much longer there are some places that you cannot go there are some things that you really cannot do where this will change my life is to give me the independence and the flexibility to go the places I both want to go and need to go when I need to do those things I give a lot of speeches about the big vision and I give very few speeches about the way to get there and I look back in preparation of this presentation as to what my life looked in the last decade in this field and I found my life is very binary they were actually moments where I made amazing progress and moments where I made no progress at all and they come almost by the year and it's really sad to admit this in public now of course the award is secured ii talked about it but it's interesting i can correlate it with some very simple feature of these projects that make all the difference and there goes about as follows in the first beginning we argued what to do and then DARPA came along gave us a goal and said do this we just did it okay then we argued more for an entire year and instead we made no progress whatsoever and then the urban challenge came along another ways and we just did it then we didn't do anything for a while and then in 2009 who will claim along and said let's pick a thousand miles just do it as almost a beauty and just knowing where you're going and I see so many teams getting stuck in the discussion where to go there can be good in the beginning but once it said this video to maintain ship for me it's a factor of 10 and execution speed just to set that direction very crispy very clearly let me say a few words about another project glass which I'm having my head wrap around their time when the South lemon car became public Larry Sergey Vic Gundotra about puppets and I got to either thinking there ought to be something interesting in the mobile space we all move from mainframe computers - desktop PC all the way to tablets and and and phones but we felt there should be a generation of compute platform that should be more advanced than this there should be something coming afterwards in this picture came up in our mind about evolution of compute platforms on the right you see Sergey on the left you see servicer gives a grand grand grand grandfather to see if he could build something that will really enable us to be in physical contact with the world free us up but have access to all the digital world at the same time and phones do a good job but taking all the fun of your pocket is a fairly tedious thing it distracts you takes about 40 seconds to even take a picture unlock the phone get the camera started so we wanted to see heavy invents and it goes beyond this there's only these kind of impossible tasks it's really hard in this jungle of technology and and and and user interfaces to find the right focus but we only made one decision we wanted to really build device that people wear all day okay so a variable device fall day now it turns out they've been people around growing devices all day for a long time you could find them at MIT exactly six people so these six people Gang of Six have been wearing devices there was just one thing wrong with these devices which is they didn't quite feel right so in the next two years we went to a whole bunch of iterations and and key credit goes to baba prophets and Isabel also no designer where we really went and tried to develop a concept in that space as logical okay we had our goal has to be variable all day this was the first version has Donna from Georgia Tech wearing a backpack weighs about 2 kilograms and then we strapped literally cellphones on our heads then we bought a 3d printer I made a little bit smaller it's 150 grams until finally the the concept emerged that we have to the present day there's a showerhead design idea is a Dalton and she designed this device over here it's a full PC dual core Wi-Fi Bluetooth battery included display camera head tracking unit and GPS lots of stuff in there that's basically Google glass as you know today and it's just been hitting the market some of you might be lucky enough to be among the chosen few that responded to if I had glass hashtag now interesting thing from the post from the perspective execution was it was an insanely logical process and the logic started out with the ideas should be worn all day and once he realizes we've won all day there's a whole bunch of things you can do it can't be heavy it can't have a wire so to make it lightweight if you make a lightweight you realize the only one display not to this place - this little too heavy this one display you're gonna wear all day you better move it above your visual field as opposed to into your visual field otherwise you're going to be distracted all the time it's gonna feel annoying and once you've done this you've realized something above your visual field it has to be a small feat of you because otherwise your eyes hurt when you're using it and the most important buddy had to make it valuable and find applications that really carry you through the day and the idea of a camera an experience gem was born or this desire because augmented reality able to see other technology that looks appealing is something we likely won't do all day long and they were likely not be as much of a reason to wear it so we focus on experience sharing and notification and so on the reason why I want to say this is this is one of the people that influenced me the most Elon Musk can have an enormous respect for Iran Iran very clearly articulated what I believe to be the cases about thinking and logic in decision-making he contrasts two different things a reasoning by analogy and reasoning by logic and reasoning a logic is the first principles and he finds that a lot of people using my analogy it's been always this way it has to be this way why we doing this for this way about always been this way most of us reason by analogy and it's very hard to reason by logic but once you're reasonably logical first principles you've either totally different solutions and glass certainly being one of those some some of different solutions from how small resolutions that's something I carry with me every single day every single day I strive to think logically ask what's really happening in society what I really the problems and then derive my decisions from that I don't always succeed but that's what I aspire to do these days I work on something new on education and I'm not going to talk too much about it I feel that higher education is another big figure that needs change just like transportation and mobile devices it's being organized unofficially it's become way too expensive it's not easily accessible it's hidden behind brick-and-mortar universities and we started working on a company called Udacity which someone you might know that puts massive open online classes thank you online our first class had 160,000 students our introduced CS computer science class just surpassed 300,000 students which means this class last year taught more students computer science and our professors in the world combined and it's a journey again which I think we can only solve with very crisp logical thinking in very clear objectives and goals and I've put together a nice team that's really pushing forward trying to develop the next generation of learning learning tools and assessment tools and who knows maybe two or three years from now I've been the luckiest iteration to report we've changed all of education the way we're trying to change all of transportation and hopefully at some point mobile devices now in concluding my short remarks at one point out some of my basic thinking about innovation and this picture summarizes better than anything else I can see this is a mountain scene I actually happen to like mountain hiking and climbing in backcountry skiing and the process for me to do something interesting is very much like the process of climbing a mountain some of us start companies to get rich you don't climb the mountain to spend much time at the top you climb a mountain because you like the process of climbing so when you climb a mountain for innovation you should pick a mountain that you like climbing not just the mountain where you like to be on top of when you pick a mountain something you want to do you can be stuck with this for next couple of years so you better pick when you really really like okay now in innovation land every mountains previously never been climbed before so there's no known route so you have to make up the road as you go really important to me is that you pick milestones along the way that are absolutely clear milestones and you agree with everybody on the moment you have that agreement life becomes much easier I've seen a lot of entities where half of the team believes they're climbing this mountain the other half be climbing this mountain is extremely painful to do this once you've done this you need a team that trusts each other even when you begin climbing the mountain the meta might be perfect you will come into a ferocious storm and you have to survive the storm and Trust to me has always been a bi-directional thing it's not just that people trust me because I'm the CEO they trust me because I serve them and I help them and I empower them and Trust has always earned it's never bestowed so Trust is opposed you have to build up for me building up trust samman team to me and for me for my team members is one of the most important things I do always do so in the fifth minute storm hits I can go on and Trust blindly my team that did us the right thing and then the thing not to do is to clear about the path up when you climb a mountain again a red line before you cannot predict what the path is you might not even see the summit at all times you have to keep the summit in mind and be excited but you can only decide what to do next tomorrow morning that's the only the size of how are you have so you have to work from day to day today and react to things and you might climb four summons and have to turn around a little bit and that's okay that's your fast failure if you spend too much time thinking about which path to go then you burn up all your energy we'll never agree and it's going to be disaster so all my projects are of the type where I pick them out and be it a thousand miles driving in California I'll be the head mounted device you can wear all day or putting an entire new allocation system in place and that summit is unmovable but the path is up to Daly surprise so this is one of my few speeches I ever given about execution but these simple principles of logic clear milestones and clear vision willingness to experiment really fast and fail fast and celebrate failures and then have a team that you really trust those are the months that work for me and that can be applied to any perspective of life it could be applied to small things I don't know how to clean my car or could be applied to big things how to change society generally have a tendency to put my time on to Hajj in society they have to keep my car you got to clean my car occasionally and-and-and when you do this I think with this recipe pretty much any problem can be solved all of you can solve our problems thank you you get stools thanks I have a million questions but I'll ask a few so what what are the criteria for choosing which projects who X will focus on with all that brainpower you have the channel how do you know which mountain to climb there's a very small number of catelleya number one you need a moonshot he suddenly looks insanely hard technologically hard so revealing into technologies and then number two is we need a team that we believe can execute and the team we typically recruit at the time the moonshot is being defined and to believe we find a single leader in Google glass of us about puppets from University of Washington and then we hire the leader together with all his friends so there's a trust relationship and when that is done we tend to set Imai someone say this is the direction you go and then leave them alone there's nothing that engineers want more than autonomy there's nothing more Provost killing the micromanagement on the top so I always found just leave them alone I can go on vacation and have a good life and still make progress now you have a lot of resources of your disposal and a ton of creativity in the team and probably you know throughout Google how do you how do you keep people focused I think the the depth of of many proteases just too many ideas I mean if you look at how many ideas really make a good good product it might be two or three and if you look at how many ideas engineers produces two or three a week so so most of his exercises are saying no to ideas and the nice thing if you if you climb a mountain as a group of people and you either want to go to the summit and you're being reminded every day every step by nature by the terrain what the real problems are and you keep say launch a product keep iterating then there's actually no space for ideas that defocus you everything is derived from extra problems in the field if you keep that loop going then every idea that comes along with your good idea okay so how about there are a lot of folks who spoke today and yesterday you talked about accountability and they talked are you looking at answers right now I great yep all right so a lot of folks talked about accountability the importance of what how are you distracted a lot of folks talk about accountability and how they shared ideas prematurely as a good way of getting feedback and people holding them to what they were working on how quickly do you share what Google X is working on whether it's internally within the company employees and more broadly and do you want to do quickly or do you wait what's your thinking around that my general attitude is to share as broadly as we can early on it was a speaker this morning very nicely said look all this kind of secrecy when it sets you back because you need other people's input to to improve your ideas now Google is big enough a place where there's enough smart engineers they can really get traction for an idea before you go to the public for some reason Google are the public likes to pick up on Google and report what they're doing so we're doing so we kind of keep it a little quiet and we love to get the moment where we had the perfect discussion when the product is basically ready or closed already but internally there's a huge numbers of discussions and they're always great one of the key recipes however is to find the day when you stop the discussion so when the portrait starts and we said a milestone we tend to stop the discussion so that people can execute now how about you know I think about Google and I think about 90 plus percent of the revenues of the company coming from the search business and then I think about okay driverless cars Google glass the things are you're focused on how do you get the rest of the company to rally around something that is just not directly related to revenue at least at least yeah not yet I have yet to meet a single Googler who would say what you guys are doing is is not good because the lack of a revenue relationship Google is an amazing company Google is all these companies every company where the DNA of the company is really given by the founders by by a surgeon Larry and and both of them are extremely curious and deep thinkers who really want to propel Society both of them are waiting burning to do more than just one thing and they have proven this in the past with Android and other products and Gmail where they've been very successful but they're now in the position that they can beat what they have and leverage into really massively new innovations now self-driving car would argue is related to revenue not living today but to revenue tomorrow and if you look at companies that don't make it over the decades and there's very few companies and technology that I say 6070 years old all those companies to make it reinvent themselves and they need to reinvent themselves because every market you have is on your market for a certain amount of time Google understands is unbelievably well I can tell you the push for Google X came straight from Eric Larry and Sergey and they've been the most strong proponents and and the Google Board has very strongly endorsed what's happening the money spent is certainly an investment it gives so many positive aspirations to the world that Google is doing but we're really serious about trying to harvesters back into into revenue in the future and so do you think that the case has to be made then that there will be revenue in order to pursue Tony's progress or do you think that is not even necessary within that culture I never really argue that I I think we great to solve a societal problem and if you if you have a big impact in society you will eventually also make some money I mean it seems to be logical to me if you don't make money you probably have my impact so but it's not it's not on the on the thinking so we never discussed that so a lot of these ideas certainly first seemed crazy and some people might still think they're crazy are we run aware these glasses around how do you how do you deal with doubters how and your team especially that probably saw a lot of the press saying oh you know some people at first said no I would never wear that do you how did how does one deal with that boy that's crazy I mean the fact that we're sitting here is crazy my voice could carry the room is crazy if you play Microsoft Kinect it's crazy you can wave your hand and dust something it's like everything is crazy almost anything you do today like I have a five-year-old tried was able to go to to operate an iPod before he was able to walk Ryan then when he puts in the magazine and does this and doesn't broke through the magazine is broken like what are you talking about this is I mean we're living in this amazing world where all the loss seemed to change and it's just amazing I mean flying cars what have you all the stuff is feasible in fact it makes sense when people doubt there's different types of doubters right as I come from a country in Germany where lots of you are a doubtful in general maybe and there's maybe the reason about to lived anymore but but sometimes people doubt because they have technical reasons and they have an argument right and some of the doubters because they can't imagine that there were too much in analogy like we've if then or argue culture by hand yesterday so we have to do it the same tomorrow I can tell you an education I'm full of doubters that tell me the the one of uncountably the professor is so absolutely essential if no data you just believe what's happened the last month ousand us to have next 1,000 years but that's the nice thing about Silicon Valley think logical whatever these people that doubt I'm happy to have a drink with them and not a second drink I guess they come back furiously doesn't say oh my god what happened okay so another in on this question what is you know is there a secret sauce you know to that distinguishes great inventors whether it's in the mindset or whether it's in the work ethic what would you say is one of the things that you've observed amongst the other folks that you know that are serial offenders oh my god I I don't think of myself as an inventor I think that almost any idea that I have was be hit by somebody else my team in my environment and it just listened and copied it I stole it this because of Steve Jobs I used to say I think it's much more for me about execution about streamlining about getting things done almost all ideas are they at a sweet truth or go to the science fiction movies you find a lot of stuff there in fact there's way too many ideas to work on at the same time to be honest it's actually the main problem here I think the the 99% quote is right on target it's it's really about how to organize things and I think there's simple recipes to to make things happen that we focus a team and that executed the maximum velocity so I don't know I feel maybe I didn't win the award I rightfully but I don't really invent I just make in a word yeah and you executed so keep on executing and smashing thank you so much for being here you

Channel: 99U

Views: 5,894

Keywords: Making Ideas Happen, 99U, 99U conference, 99 percent conference, 99% conference, entrepreneurship, Behance, design, 2013 99U Conference

Id: gLGXVXHQbn0

Length: 30min 50sec (1850 seconds)

Published: Tue Dec 03 2013