The Innovators Forum: Vinod Khosla

I'm mark Hamlet I'm provost here at Carnegie Mellon and it's a delight to welcome you to I think this is our fourth innovators forum over the past two or three years and these are special occasions for us here at Carnegie Mellon where we get the opportunity to get to talk to and hear from some of the most creative entrepreneurs of our time and today we're very very privileged to have Vinod Khosla who is an extraordinarily accomplished creative entrepreneur and a company creator and venture capitalists and person impacting the world in extraordinary ways I know you know a lot about the nodes accomplishments but let me just remind you that among the many is he's a alum from Carnegie Mellon having received a master's in biomedical engineering after his act coming to the u.s. from iit delhi i believe and a founder of daisy systems and then founding CEO of Sun Microsystems which the pioneering work in all that Sun did in open systems and RISC processors and and alike one of our other alums andy bechtolsheim and a friend of a node has said of him the note is a visionary who likes to make big bets on ideas that can really change the world and we are delighted that he's with us I'm going to ask him to come up here to the stage and in fact even Odin come on up and thank you very much for being with us I'm great to be here I'm going to ask a few questions just to get the ball rolling but we will be sure to have time for you to be part of this discussion as well because the format is intentionally informal we want this to be a back and forth and but note has indicated that he's open for all questions that this is a to be an open an interesting discussion don't be polite be a little plight so what I thought I would start you know somebody mentioned or read something in the context of the way you go about in terms of picking people and into in terms of picking projects and businesses and that you have liked to say that you invest in people as much as the initial clever business idea how do you evaluate entrepreneurs who come to you and want to get your support um let me you know it's it's funny people laugh and I say don't be polite but I actually mean it it's an essential quality of entrepreneurs to to challenge things maybe even disrespect authority you know when my kids were five I sort of probably was the main guy who threw food at them because they were too well-behaved I had the reverse problem my kids would like a little too well-behaved so my daughter just had a 25th birthday and we were at a party at rest at a restaurant with all her friends and she was telling me over the weekend that my friends think you're really cool because I actually threw some food at some of her friends and these are 25 year olds and they thought it was cool I think shaking things up is an essential part of entrepreneurship in sort of all most challenging disrespecting Authority not not being too confirming our essential parts of entrepreneurship the other piece that I find is religions really important religion in whatever you want to do most people don't know this but this is an interesting fact that most of you have probably heard of the god gene that makes people more religious turns out that applies not only to religions like Christianity it applies equally to religion about entrepreneurship this book called on being certain when you are wrong why do you have such strong beliefs about being certain when you're wrong the gene is common between really religious people who believe in their religion entrepreneurs who believe in their business plan it's the same gene written by a neuro scientist I forget his name really good book so passionate about what you want to do is an essential part to me about being a good entrepreneur being a little bit of a troll a little bit of a disrespect I have a lot of disrespect for authority and conventional wisdom and by the way for expert opinion we I'd love to talk about expert opinion you didn't ask me but what do you think about expert opinion good question turns out one of the reasons to be disrespectful for authority and conventional wisdom is exported doesn't isn't very good first I'll give you anecdotal evidence Martin Luther King said human salvation lies in the hands of the socially maladjusted George Bernard Shaw said something similar he said I am trying to remember the exact quote that but he'll come to me but this most rigorous study of expert opinion and there's obviously lots of knowledgeable professors who have expert opinions in their domains this is a bad thing to say at a place like Syria but Professor tetlock at UC Berkeley formally studied expert opinion and he took 28,000 forecasts and across a number of domains from reading brain imaging scans to what will happen in China or what will happen to the Soviet Union to how will GDP growth happen what will it be next quarter and he found across 28,000 forecasts a 20-year study about 250 experts followed closely that the rough accuracy was the same as dart-throwing monkeys for those of you statistically inclined read his book called expert political judgment those of you want just a fun sunday read the same data is captured in a book called future Babel and future Babel is exactly the essence of entrepreneurship if what expert opinion is is Babel which I truly believe and calm ox said when the Train of history hits a curve the intellectuals fall off see a lot of intellectuals among the experts in that universities not a fan of Karl Marx but true fan of capitalism but I think there's real truth so that's my view of expert opinion if we can be so wrong and take a rigorous i'll give you my favorite example a rigorous thing like medical studies on whether you should take lipitor or Staten if you have if you want to lower your cholesterol and does it impact heart disease professor iron it is who's that the Stanford School of Medicine I followed him for years when he was in Greece made the following statement he's the world's expert on medical studies so he studies medical studies he said any given study that is accepted today is more likely to be wrong than right now think about that all of medicine is based on these studies and most expert opinion is wrong you know go to your family doctor if you have little kids and they have very high temperature ask them if you should what you should do what the first thing they say take some tylenol aspirin reduce the fever turns out that piece of expert opinion has never been tested except once and the one time it was tested was tested in I ICU unit of the hospital at University of Miami among eighty some patients half of them controls and the death rate among kids in the ICU where they were doing the practice of reducing fever was so high they discontinued the study on ethical grounds so a simple fact like doctors telling you reduce the temperature can be wrong and that's why dr. ayan it is and so back full circle if you're not to know you want innovation ignore everything you know about what's expected to happen sounds like good advice I guess I know you have had such a passion and commitment and engagement in energy and environment and I want to kind of raise one or two questions to take take advantage of all all that you've done in your thinking in the area you know there's a particularly in this area in western Pennsylvania with shale gas and so forth and the predictions by some experts on the possibility of the u.s. becoming energy sufficient or whatever I guess my question is whether or not that notion of us energy sufficiency is even the right way to think about energy policy and what you think lies ahead maybe a little bit unfair because we're sort of in the midst of that shale area but so let me go back to this issue of expert opinion I'm there is clearly these experts who have opinions but if you go back to 2008 not very long ago five years ago summer of two thousand eight you asked any expert five years ago would say there's going to be a shortage of natural gas in the United States the one the largest areas of investment in natural gas was building LNG terminals to import the actual ganas five years ago and so in a typical lng project it takes somewhere between seven to ten years to build it opinion can change you in five years none of those experts are now accountable mm-hmm but they wrote their papers moved on the but they were dramatically wrong in predicting natural gas trajectory I contend it it's the same is true going forward five years from now if one technology showed up conversion of natural gas to liquid fuels one little technology all those assumptions would be wrong again right so the point is technology changes cause economic disruption and you can do what the experts do is what Karl Marx said extrapolate the past right and when that train of history hits a curve there's a change in technology all those assumptions are wrong so look I can be a better forecasting expert with about 75 to 80 percent accuracy in most places by saying the weather tomorrow is going to be the same as the weather today so extrapolating the past works the question is opinions only matter if they change what a simplistic model can do and there's been enough studies to say most expert opinion can be improved upon by building the sim most simplistic of models like in weather forecasting so you can extrapolate the past shale gas is an example or what I like to say is you can invent the future you want and that's what in a waiter's do forget what others believe I'm going to change the planet with some innovation and change the course so if you came up with one little technology to take natural gas and turn it into liquid fuels interchangeable with oil suddenly natural gas everywhere in the world would be priced like Oilers and be an international commodity and you wouldn't be this is a u.s. solution and that's a solution in Europe or Russia you'd say it's a international commodity I think it's very likely to happen in fact we have one company doing just that take the abundant shale gas and converted into liquid fuels to year three years ago they're working on converting biomass into liquid fuels and they said lot easier and a lot cheaper to take natural gas and converted liquid fuels and suddenly five years from now the experts will be saying something else I'll make a general comment about this used to be in a lot of CMU people are engineers in engineers in engineering we were taught to optimize something optimize costs optimize performance more throughput more transactions per second in a database lower cost engines more efficient engines I actually think what needs to be done where education needs to change for engineers and scientists is too learn to optimize for agility and adaptability the world is changing faster and faster at some point you say there's no way to predict what the world will do and I always assume I have no way of predicting where the world will go or just going to be very adaptable and agile which brings me to this related topic I hate titles like energy visionary I think the only thing I'm visionary about is knowing I don't have a clue I think we're where other people make mistakes is to pretend they know where we're going but in almost every area most of the really interesting things happen because you you discover it you invent it you make it happen not because it was predicted and so the notion of vision is really a wrong notion I like to say I screw up more often and faster than most people I fail more than most people and I phenomena say my willingness to fail is the only thing that gives me an ability to succeed think about it unless I was willing to fail I wouldn't take any risks then I'd be in the realm of extrapolate the past do what experts tell you what what's happening next my willingness to try high risk things is what gives me the ability to unusual things because I'm trying them you know it's sort of called you can fail and try or fail to try right pic and most people fail to try instead of trying and failing I like trying and failing I like to say I don't mind failing in fact people criticize me for saying I'll invest in projects that have a ninety percent chance of failure but use the probability of failure so much that the consequences of success are inconsequential you can take a much more in the Nassim Taleb camp as a financial investor by an option on something that's at large impact and small costs to failing try lots of little things small cost to failing high-upside when you succeed that's not about vision it's about premature confidence maybe arrogance it's about the things that are key to innovation and innovators so we're in a framework because you set forth where it's very hard to predict what's going to come to be the willingness to take risks and to explore but on the other hand if you're say President Obama or if you're at a national policy level the policies that go into place something has to sort of be chosen I know that you have had many opportunities on the national stage I recall you were here a couple years ago with President Obama in fact discussing some of these issues included including energy issues in terms of national policy takes something like like a carbon tax for instance there you don't know what it's going to lead specifically but but but you are needing to set a policy and moreover you probably want to make that policy whatever it is as clear and its precise so people sort of sort of have a sense of what of what they're dealing with what would your thoughts be with regard either specifically to something like a carbon tax or cap and trade as the case may be or other areas of that let me answer that question more generally about 18 months ago I was invited to speak to then by the Secretary of the Navy to talk to the Navy energy is a major global issue energy is a major driver off the Navy and what they can can't do you in the middle of the ocean and can't get your fuel your you got a problem so they asked me there's a hundred roughly 150 Admirals in the Navy they were having a meeting planning for Navy 2040 mm-hmm and they asked me to open this sort of conference they were having for just the Admirals now think about the problem you have to plan the Navy you can't build aircraft carrier in less than 15 years you if you have research it's even longer to first design new capability like for the new Joint Strike Force fighter so what do you do huh and but I started by asking this was 2011 I said in the year 2000 what was the Navy's mission top three missions I turned out China didn't figure in the mission we forget the China was not politically considered important Japan was the other economic powerhouse and not that long ago and I said if your mission has changed but there was no 911 there was no al Qaeda there was a very different world for the Navy and I said if you didn't know what your mission was in 2004 what would be important in 2011-2012 how can you plan 20 40 and they now have a new set of missions I call that the pretense of knowing you pretend you know something that you known no you're much so the question was then after I gave my talk about this baloney about planning the Navy for 20 41 dead medals ask me so what should we do my answer was if you know you don't know you do different things than pretending here's the three scenarios you have to plan for and their response was the world's changing really rapidly instead of three scenarios let's plan the Navy for five scenarios and I said if you did 15 you still wouldn't predict Osama bin Laden and one guy causing that much change in global politics and if you suddenly discovered shale gas and energy wasn't a big issue suddenly your mission changes it's no longer protecting the sea lions by the way just protecting the oil lanes in the last 25 30 years we've spent seven trillion dollars which is really a subsidy to the oil industry because we're doing their transportation insurance for them so when you think about it this way you engineer these systems to be adaptive and agile which is a very different cast than taking any mission whether it's one or five and optimizing for them and Don Rumsfeld after fighting the Iraq war essentially said we no longer need traditional defense which upset a lot of Defense people we need agility for the unknown unknown I think that was his famous line I think what are you building an engineering system a computer system I mean imagine computing workloads have changed I know the head of the computer science departments here workloads have changed so dramatically Big Data Hadoop all that not one piece of architectural change has happened to accommodate those and I can't find a place during computer architecture for these new loads and so somebody asked me so what would you do and they sort of said here's the new architecture I said no wrong approach that's new architecture if you assume Hadoop is your big data computing environment and big data revolves around her do but almost certainly in five years nobody will be talking about Hadoop as a big data computing environment let's build an adaptable system I said why don't we instrument the device to reconfigure some things every minute some things every our some things every day and some things once a month huh now nobody's doing that in computer science that I know of that I've been able to find so adaptability and agility in the face of rate of change become the most important things to optimize for make sense I want to take advantage of being the moderator to ask one more question but then I'm going to also hope they'll be to keep hogging all the questions here and turn it over to some all the guy talks too long no no no no no I but I do want to get one one other one in so we're very excited we are opening up a major energy institute thus God Institute I know you've been involved with some of the discussions that we've had the opportunity to have as we think through in fact we're even more pleased we receive not only the generous gift from from the Scots but 30 million dollars just this past week we were able to announce in the Richard King mellon Foundation I know that one of the things that we would say that we are particularly focused on as an area where we think we can bring something to the debates and policies on energy and an environment has to do a sort of systems of systems sort of systems analysis but probably taking one step up and I know you know a lot about what we sort of have as a backbone in that in that regard but would you have any advice for us it's kind of a selfish CMU central question but do you have any advice for us on how you think we should prioritize you know we're an academic setting this new energy Institute's not that we don't have lots of things now but now that we're able to kind of contemplate putting them into into a package what are things that would come to you the things that come to mind are generally the plans I've seen are two linear and the world is a nonlinear place at the one end the chemists are doing chemistry in the chemical engineers doing chemical engineering and the physicists are doing their thing diverse interacting systems result in much better outcomes to the point where about 10 12 years ago when I took a sabbatical from investing I sent it spent it became a postdoc at the Santa Fe Institute in complex systems theory my kids loved it by the way because I was sold on my math had to get a math PhD from Stanford to come to to me for six months to brush up my math so I could actually go build complex systems models I think complexity isn't included enough and nonlinear effects from comp come from complex interactions I mentioned medical studies one of the causal reasons they are wrong is if I am a drug company and i'm doing a diabetic drug right i really don't care about improving the world i care about getting my drug approved so what do i do I get a population that's high likelihood do not have complexity in it so while clinical trials so hard because they select somebody who has high blood sugar but doesn't have hypertension they don't have cardiac disease they might not have thyroid disease turns out they might not have mental illness turns out depression and diabetes of comorbidities so they get very high approval rates because this person is taking one drug only theirs and the trial passes in real life most of the people who have diabetes taking a drug for depression for cardiac disease for hypertension and complex interactions start to happen that's why most studies are more likely to be wrong than right which it just boggles my mind that we don't do something about this fundamental fact every piece of medicine with every pill we take is based on studies that are more likely to be wrong than right you just makes me shiver so complex systems are really important this morning we were talking about electric car batteries a little bit and I said do you realize electric car batteries and whether they're viable or not are impacted much more not as much but much more by weather machine learning and robotic autonomous drivers and driverless cars become popular then by improvements in lithium-ion chemistry I will challenge anybody to challenge me on the following fact any improvement in lithium-ion chemistry which is the assumption on how batteries will go and under another assumption of experts I don't believe almost certainly the solution will not be lithium ion in my view but any improvement in chemistry in lithium ion is likely to be less consequential to electric cars than autonomous driverless cars why for the simple reason if you had an autonomous car you could get a car to drive 10 times as much per year if you can amortize that 10 times faster you can't get a 10x improvement in lithium ion economics you can get a 10x improvement in autonomous vehicles driving the car so you get a hundred thousand miles per year instead of ten thousand miles per year per car and acid used that's what i mean by complex system so i think that's very important at the other extreme almost all the efforts I've seen I've seldom seen computational physics computational biology computational material science computational fluid dynamics used enough so interactions across disciplines and more fun science is key to energy than incremental improvements in what we know so those are two opposite ends one is complex systems interactions and the other is really more fundamental science very everything in the metal is mostly inconsequential ver very good advice now we're sitting here and there's all these lights and so I don't know that see out quite so well but I want to turn it over to questions from the audience and if I kind of squint it you'll know why so if you have a question just raise your hand i do see okay right right there hi so you don't say also who you are and where you're sure i'm a senior math major um Elliot given given your conjecture that you should challenge the status quo very often how does that affect how industry and academia should interact in the context to entrepreneurship right so if you're an entrepreneur and you're interested in working with a professor or you're a academic and you're interested in taking something that you've discovered and bring it into industry how exactly given that you are sort of the status quo how does that change no you know the same polym would say don't go to college alright his his book new book aunty fragile is way more important than most people realize this some fundamental ideas and they speak it's called anti fragile for a reason he argues that we build overly engineered systems and what we're really doing is increasing fragility and in error prone systems you get more robustness more maturity and it doesn't matter whether it's in financial systems in the financial crisis will and he essentially talked about it happening before it happened so he's one of the few people I respect you build robustness into the systems he would argue that doing is more important than learning at least credentialing I'm not quite a believer in that but I do believe in mixing learning and doing at the same time is much more important and it changes what you learn and what you retain and how you learn I'll give you an example in our own field there was a guy who used to work for me and we have a rule when young guys join our venture firm khosla ventures they cannot stay there too long that means there's no career path to become an investor by joining us which is pretty controversial but he was with us for three years and then he did his own startup and started things and he said to me a year later he said I heard you speak advise entrepreneurs for three years he said I interpreted them very differently then I do now now that I'm an entrepreneur so the same teaching but you know in teaching it's not what's the message that sent that's important it's the message that you received we assume message sent is message received but that's not true mostly learning is worse than the game of telephone whisper telephone message received is seldom like the message sent and the message received can change based on what you're doing how you are interacting how much in real life so I think interaction is much more important I'll try and keep my answers shorter but I I think it's pretty important it gets very relevant for what we strive to and hopefully can strive perhaps even further and trying to make real world problem solving and learning and doing and making kind of part of the part of you look in some departments it's much easier to do in computer science it's much easier to hack together a website and and my son's our sophomore in computer science at Stanford then you're much easier my daughter's a junior in civil engineering it's much more much harder to be as hands-on right and interactive so it does depend my oldest daughter design which is a mechanical engineering discipline at Stanford so you have to meet and there was easy to be hands-on you could study about materials but when you did design and then tried material science and need to know about properties of materials your view of Port de Belon changed in material science changed we I think we got a good sense walking in through the hull here of other the hands-on I think in the west bend the art students yes uh-huh but by the way I being in the cfi I'm against liberal arts education in undergraduate I think it's the worst thing you can do first every engineer should get enough liberal arts education but the flip is true I wouldn't consider anybody educated if they didn't have any computer science science Technic engineering education you are not our view of liberal education is what came over from England in the 1800s it's silly to think that a liberal science major doesn't need to know computer sign we have this silly language requirement and high schools and colleges the most important language is programming not French if I had to pick between the two which one would you pick my I was proud of my daughter's change ferd has a language requirement sure she decided to use sign language which is spend the least time wasting least time wasting your time learning something that's a historical anomaly so I think now if you want to stand study French literature only time or fine arts or the cello only time it's important is in a graduate program right basic education should be the same for everybody our society has become so knowledge-based that we should stop extending the past you know in the past we said everybody should get a high school education and then do what they think well that benchmark has since moved to basic education where you understand how your car works how engines work as much as how Shakespeare works in fact I'd venture to say Shakespeare is not as important and knowing how your thermostat works which I find ninety-five percent of people on the planet still don't know how to operate a simple thermostat intelligently on on their wall so the nature of Education has to change what's essential nobody should graduate from CMU without being able to read every article in The Economist every week you can't do social science and psychology till you know statistics otherwise you get stupid social science which most of social science is things that are not validated or rally dateable or provable so if you don't know statistics don't tell me you're doing psychology I don't want to talk to you if you know statistics then I'm really interested in what you can tell me about behavioral science in fact there's no engineer who should graduate without multiple courses in human behavior in psychology or economics really important but based on real provable repeatable facts sorry that's a beef of mine whoever's from Modern Languages we're going to add Java and C++ s well so so you bring up an important issue you shouldn't be able to do language without doing linguistics absolutely I mean linguistics should be an essential part of every curriculum whether you're doing material science or you're doing music or the Curie of music should be really an essential to me those are the essentials of learning how to learn and think in capabilities actually I think that you know here are like our language tech nology institute which is basically a department in computer science but it had its roots in linguistics philosophy and hillis tree should be essential linguistic should be essential but not the easy parts of those hmm right literature you can do all day long on your own if you're smart educated person Jane Austen is not as important as most of the things the structure of language the structure of music all really important so people over interpret the traditional notion of liberal arts should be abolished and anybody wasting 250 grand of their parents money to get a degree like that and then be unemployable I got into a controversy recently because there was a guy who did a Yale program and then tweeted the only job I can get is bartending and I said that's courses you took and wasted the privilege of being able to go to Yale and spend 250 grand of your parents money and you forgot the obligation you had to use a 12 do you know this movie men in black where you take the little thing go and you'll forget all this on any way out the dark when you you won't remember any of what what we just what was just discussed um let's see it was such actors like there's little i have strong views on every yeah i think you that was where i left legs yeah i forgot my question I'm only kidding I'm Alan rosenblum a physician researcher former life I was in ICU and I never treated infectious fever there it's 5,000 years ago the Chinese have a symbol for fever blah blah blah Blee know it clears infection anyway my question is navigating the minefield of this yeah by geography navigating the mine field of entrepreneurship there's a whole bunch of icebergs and one that really disturbs me is given I'll give you two short examples one I learned over here at the extra electric garage somewhere around 90 for Toyota had a electric car the battery the guy who invented the bed this is what they told me over there so I'm quoting them sold it to GE figuring that GE would mass produce at etc GE then looked around like any company how can we make the most money off of this well they decided to make the most money off of it by selling it to shale oil you know it happened shell suit Toyota 470 million for okay that's example1 example2 I'm personally know a group that invented a beautiful handheld of what a beautiful handheld flow cytometers and gets cd4 levels very important in treating HIV and this thing you can't have a sheath flow so they use sound focusing brilliant design bought so the VCS wanted their money back and they said okay we're going to sell this off to and be trogen the scientists this is in New Mexico well let's keep the question short let me answer that anyway invitrogen bought it and killed it and that was their plan all along so how can you prevent your best intention and your great invention from just being bought and killed yeah so this is a great question for a forum on innovation a couple of things one that realities of life is not every good idea gets into the market and there's 20 the fundamental notion of innovation implies change change implies disruption and most of existing society benefits from not changing things not disrupting things I'm there's always an interest in keeping things the way they are the oil companies don't water on biofuels or electric cars that G doesn't want to interrupt this gas to combined cycle gas turbine business so there's most of the established interest is in not changing thanks there's again expert opinion that what you ought to do if you want to scale something is go to the large companies I was talking to the gentlemen in robotics doing the driverless car at CMU last night briefly got in late he was willing to meet late at night and he said you know we're working with General Motors and I said oh no almost a guarantee of failure right then first they have no interest in disrupting cars if you have a driverless car you'll need fewer of them GE won't do that General Electric one hour General Motors won't do that all right the other is they have a good business and everybody there is on a career path and career paths have asymmetric benefits if you succeed you get minor benefit if you fail it's permanently on your record so it discourages this notion i talked about failure big companies don't like to fail which means they take only incremental risks no large innovation has come from big companies did Toyota or General Motors do the electric car or did Tesla did NBC new media change or did Google did walmart do retailing change or Amazon I could go on and on and on I've really never seen a big company in a weight or be able to effectively bring innovation to market why because they over planet they think they can predict the future when all you can do is try little things fail pivot left pivot right use every failure to learn adapt and change and that's not a comfortable thing in a large company and that's why all large project spell show innovation doesn't start from big companies and they can't scale it at some point you can get them involved when the risk is low but while the risk is high no big company can do it and culturally they don't have what reward risk-taking and hence innovation big companies oxymoron together um so and and the other reality is you will have failure because people will want to kill technologies that's not the first time it's happened so unfortunate part of life is real life has some unpleasant things you know I don't like to see animal killed but it's just part of life so I'm sitting this way and I want to make sure I don't short change over here so right there is that yeah okay my name is Daniel I'm from computer science so you'll be talking a little about being afraid so I wanted to go a little deeper into this the nature of fear when you're trying to go forward with interpret like to speak up somehow the acoustics are not set up just to get sound from their here I get a design from here to there s message sent not message received this okay so you would mention fear being afraid of moving forward different things so I wanted to go deeper into this nature of fear and how to actually attack it and I wanted to go at it at a specific example so since you're related to energy well at least it says that over there and how about let's say nuclear energy are we too afraid of it it's supposed to be the most amazing technology that I personally have seen in terms of you know being able to produce that from a very small piece of materials so what will be your thoughts on that kind of thing yeah um you know my favorite saying on fear is this you know people like to talk about courage and especially courage of your convictions I talked about religion and passion about what you do but the word courage is meaningless if you're not afraid there is no courage without fear you have no fear of something then doesn't take courage to do it it's an essential part of risk-taking if something was low res somebody would already have done if something is risky then you have to have fear and you need courage all I can say is certain people i incline that way some people aren't one of the things I mentioned the god gene earlier i think the god gene actually minimizes the perception of risk so you need less courage to do something by making you a little bit stupider about the real risks and i often say if i detect if i knew all the risks in a start-up before i started I'd mostly not started all right and so this sort of blissful ignorance premature confidence arrogance all these essential ingredients of innovation and entrepreneurship as well as sort of overcoming your fears the best way i've done it is i start in life not needing much and i said so i have very little to lose and a lot to gain because if you don't have much you don't you can't lose much in fact when i did my first startup i said to one of my VC friends i think this country is great it's a great deal where if we win we went every lose you lose because i'm not putting any money essentially the only thing i had to lose when i did my first startup was my student loans there was nothing else I could learn and so one phrasing it right and saying what is the risk you're taking you might risk take a risk at a technology development and it may fail but if you view it as failure that's one way if you r you it at as learning with somebody else's money that's pretty good you don't have to spend 250 k of your parents money spend some venture capitalists money learning and then if you sort of say I'm better off in doing my next thing it becomes much easier to take the risk very good let me see right there if somebody can see where I'm pointing to see hands I don't see people but there's the microphone yeah okay hi my name is Matt I'm a first year MBA student my question is entrepreneur's spend so much of their time championing their ideas and convincing the world that they have the next big thing I was curious but before it is I was curious if you could maybe share your story of Sun Microsystems of that moment where you knew this was going to be the next big thing I like to say that the huge difference and startups often confuse this between strategy and tactics being obstinate about your vision is important you're going to go change the world and that son we knew we wanted distributed computing to happen and everybody told me why don't you become a graphics terminal on deck back says deck has disappeared but that was the dominant computing environment deck an IBM and building graphics terminals for them that was the incremental approach right but we were very very clear we want to do distributed computing and we were i would say foolishly visionary first we made networking Ethernet standard on every sun we ever ship nobody had ever done that in computing but right but his his the scary part we wasted a lot of money every son shipped early on had DES encryption because I imagined the security world as being important to networked in distributed computing it was a bad idea because the world wasn't ready for security so as though my strategic vision of needing DES encryption and every node was actually exactly right in 1982 tactically implementing it wasn't so it's important to be obstinate about your vision and strategy but you have to be really flexible about your tactics and not get stuck up usually people who are visionary will not compromise that vision to be tactically effective and the way to think about data is you can't get to your vision in one step you need 20 steps along the way and at each of these steps you garner some resources nice people believe you so you can do the next incremental thing so you pretend to do little things and sometimes of course things like we had a huge fight and somebody eventually convinced me to take networking out of the Sun and put what would then call me t100 rs-232 ports so it could be used a time-sharing machine that was that tactical compromise of my vision but it got us enough revenue to get the next funding to do the next thing and so being very very flexible on tactics but being obstinate about your vision is an important part and that happened many many times its Sun and that's why I say you discover your business plan you don't make it but your vision you should be generally something you're passionate about happening and you should be flexible about your vision a little bit but but generally if you're too flexible with your vision you become to opportunistic and give up on the larger picture great oh yes right here yeah go ahead I think the microphones coming to you my name is harsh monthly can you hear me yes and I've just taken the leap from the corporate world to being a student again as an adjunct faculty my question has to do with coal energy and India and China you made a very compelling presented a very compelling and stimulating point of view on trying to invent your own future India and China are mushrooming or the use of coal and the result in pollution is mushrooming what approach would you take if you had that problem well I'll tell you what we attempting turns out when your economy is growing at ten percent or six percent or eight percent a year its infrastructure is generally growing fat needs to grow faster that means roads buildings construction so I personally taken the approach and i'm often wrong so let me start there is to say all the expert opinion about taking co2 separating it and sequestering it is complete baloney it is exactly the wrong thing to do but if I can use the carbon dioxide effectively carbon dioxide is bad calcium carbonate which is probably this stone here is one form of carbonate is good so with the magic of electrochemistry if you can turn co2 into co 3 carbonate ions then you actually have really valuable products we just built a new building fifteen percent of the cement and our floors and walkways is all carbon captured as cement so that's an example that requires hard innovation in electrochemistry to make the co 2 2 co 3 conversion cheap and that generally means lowest energetic cost of producing a no h ion i won't go too much into the detail at the other end it's taking carbonates and doing finding nano structures that increase their value whether it says granite or marble or cement or something else so that when you ask fundamental questions as opposed to read expert opinion and believe it you come up with very different answers I find that's an example Cole would then become economic because you'd be co-producing cement in those economy is needed I'm looking at my handler who is telling me for reasons I will explain that in a second that we we must wrap up here in just one more question I believe so i have a concluding item after this question but let me pick somebody for the last for the last question that hand right there's been the most vociferous hey my name is magan i'm a chemical engineering major at carnegie mellon yeah i'll speak up back up oh I'm a chemical engineering major that was it ah and so you talked a lot about yeah not being certain and not and knowing that we don't know that much what do you think in general about having the control about future events and is that even the notion that's like desirable having the control about what happens yeah so the question i'll repeat as I understood it is you have to control future events what do I think about it obviously that's true at many many levels the fundamental thing I would say is we focus too much on providing reducing failure which means increasing control and predictability there's another way to make failure more acceptable and that's to reduce the consequences of a failure so let me give you an example if I take a computer system and flip a chip or cut a wire it normally fail if I cut my finger my body keeps going it's a resilient error-prone system it's tolerant to errors and mistakes and failures by and large I think because of our engineering mindset we've done more of prevent failure stuff and less of reduce the consequences of failure so we can tolerate more failure I think at the social level especially at complex systems I think we should engineer for more failure but make it smaller that's what I think to the extent control becomes in the electrical engineering world a software problem and all software is free essentially and mostly computing is free putting more power into control is a good thing let me end by saying one generalized comment it relates to your question but in a much more generalized way the world economy is becoming more about ideas the capitalist system was invented when labor was the principal way earning happened if there's an economist that we have it they'll be happy to hear this especially liberal economists if you got paid for your labor you only had eight thousand hours of labor if you worked a hundred percent of the time in a year unfortunately or fortunately I actually consider it's fortunate the economy has moved to one driven by ideas and so labor is no longer an important component of the economy ideas are that's why innovation becomes important that's why entrepreneurship becomes important that's why I believe the power of ideas fueled by entrepreneurial energy is the single most important force in the world for good and intellectual horsepower learning real learning become really really important in the fate of the planet and you get phenomena like a mark zuckerberg can make 25 billion dollars or whatever he's made org larry page because multiplying ideas is easy multiplying a car physically a hundred times costs a hundred times as much maybe fifty times as much because of scale economics and all that but multiplying an idea doesn't take extra costs and so ideas become disproportionately powerful tools and disport portion utley more valuable so as long as the economy moves to the power of ideas driven by innovation we will see more asymmetry in economics the disparity what's called the Gini coefficient in economics between the rich and the poor will keep going up the poor won't be the people with the resources will be people with ideas so social mobility will hopefully increase but it is a fundamental change in the structure of global economics and capitalism that we haven't included in the theory of economics but innovation is the principal reason and the power of ideas intellectual ideas is what's driving this thanks thank you very much if that was really nice thank you thank you I'd like to thank our guest he has an extraordinarily busy schedule is going to be meeting with President Suresh in Philadelphia before a few a couple more hours and then I know you have speaking come set of commitments in New York City but we are speaking engagement in Philadelphia two o'clock so we are doubly grading at new york city's i apologize i have a run we are we're slowly let me just say anybody who has communicate i know people who have questions for me feel free to email me at VK at khosla ventures com I'll be impolite and run and remember being impolite is important if if people are interested in impolite I gave a talk to a high school kids a lot of Indian kids and Chinese kids in this high school and I gave my talk about why they should not listen to their teachers why they should throw food color outside the lines it's on our website I actually think it captures some really important ideas about behavior and what is acceptable and what you should do to challenge what's acceptable and let me say that for all of you who are liberal arts majors please we're gathering in the corner after this talk for the therapy sessions that will be required thank you very much thanks a lot that's great thank you

Channel: Carnegie Mellon University

Views: 9,653

Rating: 4.8823528 out of 5

Keywords:

Id: A-lkJF1WvDQ

Length: 65min 23sec (3923 seconds)

Published: Thu Apr 25 2013