Sam Altman Startup School Video | Waterloo Engineering

good afternoon good afternoon and thank you for coming to what is going to be a very special afternoon for you today we are here today to see and hear Sam Altman president of the YC group which is the parent of the world-renowned Y Combinator frankly it wouldn't matter what Sam was going to say I would want to listen but today he has promised to give us 90 minutes of startup school a program he launched earlier this year to make startup knowledge more accessible to the masses sharing knowledge that's Sam's thing let me tell you a

little bit about Sam although I'm sure most of you know a lot about Sam already he's become a legend in Silicon Valley despite his young age but we know that age is just a number is what you do with your knowledge to improve human lives that really count Sam was just 19 when he co-founded his first startup company looked which was acquired in 2012 by green dot for over 43 million dollars well they turn around and name in the chief technology officer and later to its board of directors through the next five years he co-founded

hydrogen capital served as the Acting CEO of Reddit and the chairman of the board of two nuclear companies I met Sam in 2014 and I think it's about the time that has just been named president of Y Combinator or YC perhaps the most prestigious startup accelerator in the world in 2015 Sam was named the top investor under 30 by Forbes in 2016 he co-founded open AI a non-profit research company established the sphere development of safe artificial intelligence for the benefit of humanity today Sam Altman for the most influential figures in the Silicon Valley and a

North American technology sector the startup school series sounds way of bringing essential and free startup knowledge to more people if you have an open internet connection you can learn from the experts by watching this carefully curated program the knowledge gained a start-up schools avail invaluable as it is participating as it is in participating in YC itself in fact more than a few of our Waterloo engineering students and alum have used this participation in YC as a springboard success including the founders of thalmic labs vidyard pebble and buffer box Sam has had a huge impact in

the Waterloo Region and we deeply appreciate his support I'm also delighted to share with you that Sam is receiving an honorary degree of engineering degree tomorrow at the convocation during the Faculty of engineering afternoon ceremony where he will also give an address the convocation is livestream so please watch for it he is an intellectual leader was inspired the world and inspired us we are very proud to have him as an alumnus now please join me to welcome mr. Sam Thank You Dean for that lovely introduction it's really nice to be back here I remember the

first time I came to Waterloo I felt like I really would have loved to go to school here the thing that stuck out the most is that unlike where I went to school I had to find a group of people who are interested in building and creating new technology and not just sort of going to class all day everyone here was like that and I really love that and that spirit has been amazing I think I'm pretty sure this is accurate I can't think of anything that even come close I believe there are more Y

Combinator alum companies within a kilometer of velocity than anywhere else in the world outside of the Bay Area so this is a special place to us and it's always nice to be back this also yeah this is the first time I've ever been here in this summer and that is an extra special treatment that's been great so I'm happy to be doing this talk today because this is this is the last day of startup school so we just finished we advise 3,000 startups at once online sort of a crazy experiment and worked really really well

I have given talks like this many times I have the most fun when it's mostly me answering questions so I will talk but please interrupt anytime you'd like I'm happy to go in any direction whatever is most useful I'd much rather be interrupted if I'm boring you or telling you something that is not helpful then I would to sort of waste your time on something you don't want to hear about thank you very much so this is great I had a late flight last night so so I wanted to talk about the question that I

usually get when I'm anywhere other than Silicon Valley or maybe a little bit less in Waterloo than in most places which is what makes Silicon Valley special why how can we be more like Silicon Valley what do we need to do and first of all I think the most important thing is to not feel like you are any less impressive than people in Silicon Valley I think people handicap themselves a lot by saying that well you know I'm gonna grade on a curve and I need a few extra strokes in the golf game because I

haven't I'm not Silicon Valley completely wrong obviously talent is is equally distributed around the world there may be more access to capital in Silicon Valley but everything else is basically harder it's harder to find office space it's harder to recruit and retain engineers so I think the number one piece of advice that I always give people is get out of your own head about this but but here are the things that I think are Admiral but Silicon Valley many many things we do wrong here are some of the things I think we do well and

are worth trying to emulate one is there is there is just a relentless belief in the future that there's going to be a future that things are going to change the future is going to be better that it's okay to say that the future is gonna be better it's okay to have strong ideas one of the hard things if you are trying to build a really successful company you kind of have to live in the future Alan Kay talks about how he tries to mentally live 30 years in the future but that's a little bit

too aggressive so he tries to build the future that's 15 years out 15 years is a long time technology trends are these exponential curves it's really easy to be building the future be correct but seem seem kind of very crazy very out there very focused on the wrong things in the present and one of the challenges of this is if you're going to have this relentless belief in the future if you're really going to truly think about the world thirty years out and try and then get to work on building the world that's 15 years

out people are going to mock you and all all humans basically have this evolutionary drive not to like me and may sum up not to like me and muck you know you want to fit in with your group and one of the things that really works about Silicon Valley and one of the things that works about Waterloo too is that there is enough critical mass of people they do believe in the future that you can find people who want monkey you can find people who think it's ok to be ambitious we think it's ok to

be working on the problems of tomorrow we think it's ok to be trying these things that sound like they're not going to work today close to that is just you want to have some people and I have found a group over the last decade plus of people like this for me but I I have people that I can call up you know at 11 o'clock on a Friday night from a bar with a really crazy idea that sounds totally wild and completely nonsensical and they will at least humor me they will not immediately say that's

dumb they will say well let's think about how that could be good or well that sounds really bad but maybe it's good for this reason and most of the time they're just bad ideas but every once in a while they're really really good and the problem is most people just kill all of their good ideas I don't know what the Canadian phrase for this is interestingly in the u.s. we have no such phrase but most countries do I picked one from Australia called Paul poppy syndrome but this idea where if you try to be too

ambitious people try to cut you down is there Canadian phrase for this what is it you can feel it so I think it's really interesting that there is no there's no American phrase for this that it's at least popular and it's good that there's not an obvious one in Canada either but I think it's really important to find a group of people who won't try to cut you down for saying like I'm going to go build this giant company again another really good thing about Waterloo density really matters density is really really important trying to

be around you know a lot of other people working on startups in close physical and mental proximity is really important and then the other thing that I think we have for all of the cultural problems in Silicon Valley we have a very good culture of paying it forward so people help you and that you're successful you are expected to pay it forward and that's been great to see at just among the YC companies that I know of in Waterloo that there's a real culture of that and I know a bunch of YC alums are running

many of the startup related orgs and starting a new one and that happens in Silicon Valley as well and I think that's that's a really good sign so another question is that we get all the time and that I have spent the last ten years of my life trying to figure out is what makes really great founders special there are two phrases that kind of come to mind if I were trying to if I were trying to like pick what our top ten most successful founders would have said about themselves when we were interviewing them

at YC I always figure it out and I never give up would have been high on the list and everyone thinks the thing that really matters is how smart they are or their domain expertise or their network or their connections it really is this kind of a personality trait people have different phrases for that determination relentlessly resourceful as one that Paul Graham uses but that spirit is the most important factor I think in successful founders you need to have everything else but a lot of people have everything else this I think is the kind of

the rare and the limiting trait the good news about this is I think people can kind of learn it people don't try too often but I think this is possible where if you apply yourself to getting better in this way and the way you do for anything else I think people make huge strides this is not this is not something that if you're not born with it there's no chance three three things that we have observed about how successful founders get things done focus self belief and personal connections these are this actually started some in

Charlie Rose said that he had observed that the way things people got the way people got things done in the world was a combination of focus and personal connections and then for startups I think it's really important to add this you actually have to believe in yourself you actually have to believe you might do it but these three things we've seen again and again and in fact if I were looking for a co-founder I would look for a lot of that like is this this someone have a relentless focus are they just going to get

this one thing done and keep their blinders on not get distracted by shiny objects along the way do they actually believe that this is possible because momentum is this crazily self-fulfilling prophecy and can they form the personal connections that it takes to be successful will they be able to recruit and retain a world-class team will they be able to sell the product will they be able to raise money will they be able to talk to the press the ability to form these personal connections is super important another thing that I did not used to believe

was this important and now I do is clear communication and vision and thought is part of a really clear communication almost all of the best startups that we have ever been involved with from the very first timing methods founders they were able to very concisely and clearly communicate what they were doing in like 25 words and I don't know why this is so important it may be that you need this to spread the message but I can certainly say that founders who aren't good at this don't really go on to be successful founders that don't

get good at clear concise communication really struggle and you can you can prove this to yourself quickly by just looking at the founders of really successful companies they're all good at this so I think this is an area to invest in and get better in if you have those things then one thing you can do is attract great people to work on the company with you which is hard and important it's really really hard to recruit xx employee for a start-up one thing that I always tell people is it is easier to start a hard

startup than it is to start an easy startup and the reason is you need a startup that is going to be inspiring you need to startup the people are going to want to help you with and if you are another startup that kind of fades into the noise and no one really cares that much about maybe a fine business but it's just not exciting it doesn't no one wants to contribute their energy to it it's really it's really hard recruiting that xx employee is really hard you need an exciting vision and you need to be

good at communication and personal relationships and then also founders just get a huge amount of work done themselves so in the early days especially you kind of got to do everything and there's a lot to do and so people that have this sort of like focus and maniacal productivity is really important the thing that people always want to talk to me about though is none of those things people always want to talk about the idea the idea is important for sure but there are a lot of good ideas in the world there are far fewer

people that are willing to put in the effort that it takes over years of decades to build a great company then there are people have great ideas that said starting a startup with a bad idea is also bad so I'm happy to talk about this and about how to have good ideas but but this is always what people want to talk about a lot of people think if they could just come up with a good idea then they can start a really successful company we've run this experiment at YC we funded 21 companies where the

founders seemed really good that they didn't have an idea all 21 failed in the first year by far the worst worst external we'd ever run it turns out that good founders didn't have a lot of good ideas and also good ideas without real commitment behind them are just tough so wait to have a good idea before you start a startup if you start a startup without a good idea and you just kind of cast around for one you'll be under pressure to make something up it won't work that well and startups the very best startups

are all started because someone believes so passionately in an idea and they believe that a startup is the best way to make that happen one of the things that is really hard about good ideas is that original thought is really hard most people just copy somebody else's thoughts most people do what somebody else is doing and this is disastrous because if you have one idea that starts to work and you have 10,000 clones that follow it none of you are going to be successful and the person that started first that you're copied is very likely

to win you know in the year after Facebook started there were well over a thousand social networks that started in the year after Instagram started there were well over a thousand photo sharing apps it started we have not seen the same snapchat cloning race I'm not sure why but you want to do something that is new and different you want to think of an idea that's not the same idea everybody else is working on um either year that I joined the board of Helion which is a nuclear fusion company according to somebody's count there were

10,000 photos startups of some sort or others started that year and one nuclear fusion start-up started which seems a little bit off to me but I think most people just do what everybody else is doing right then noticing problems in your own life is certainly a good way to do this it's not true that all successful startups are started to solve a problem that the founders have themselves that is a very high percentage if you really go back and think about the transformational companies most of them start with someone solving their own problem and then

there's this idea of a great wave which is that another thing you notice if you go back and look at all the really successful companies is there was this massive wave of technology coming you could see it in the distance and you knew it was going to just like crash over the shore and through the city and you got out there early on your surfboard so I think it's really hard to create a technology wave as a startup that takes too much too many resources you can't do that but what you can do is notice

that a wave is coming and university students are particularly good at this you can notice that a technological wave is coming and while other people still think it's a toy are not going to matter you can really get conviction behind it and serve somebody else's wave and this by the way I think is the reason that great company there tends to be these clusters in time when a lot of great companies started so you know when the the internet wave happened in the mid to late 90s you had a huge cluster of companies that are

now very large Amazon Google Yahoo I guess others and and then when the mobile wave happened after the iPhone came out 2007-2008 you had this other big wave of uber Airbnb snapchat Instagram whatsapp in this short period of time right after this this new wave started and so very good question think about now your Facebook I would say kind of was the one that straddled both they did really well on the internet and then even better on mobile and a really good question asked is what is the way that's starting right now and what will

be you know where is the cluster of companies because we're about due for another one where is that going to be another question that people come up with all the asked all the time is about co-founders and how to find them we do have a significant preference for co-founders I think this is borne out in the data we are very happy to fund single founder companies and I would guess you know on the order of 10% of the companies we funder single founder yes I don't answer and I'll tell you why I am equally likely

to be right as wrong but because you might listen to me I could set you down like a terrible career trajectory and I would feel very irresponsible if that happened so you should not start the startup I want to start you should start to start if you want to start and you should trust your own convictions you look younger than I am you're more likely to be right that's sort of been the history of Technology but I mean one that I am so sure I'm right about that I'm willing to say is I do think

machine learning applied to every vertical is as close to a guaranteed shot as you can have right now in startups the problem is a lot of other people think that too so it's brutally competitive but there are going to be you know 100 million dollar companies that get started around now applying a different verticals so that's one that I that's one I feel confident enough about to say but you know if you think it's Bitcoin or VR or synthetic biology then you should bet on that so you know Amazon I think was basically a solo

single founder company and obviously that's been fantastic it's clearly possible and it is certainly a lot better to be a solo founder than to have a bad co-founder I actually was dealing with this this morning this is about the time in a YC batch when people that kind of came together as cofounders just so that they could get into YC and told us how they knew each other for a long time and made all that stuff up this is where that all falls apart so you really do want shared history with a co-founder it is

a very stressful thing it's like a relationship and if you don't have some goodwill built up in the bank it's like if you start a relationship with the most stressful bad part and you need that shared history you need to feel like you owe each other something in that way at some point the the expected value of most startups kind of dips below the x-axis and if the co-founders feel like they're friends and they owe h other something you can kind of get through it as I mentioned earlier this concept of determination hugely important and

one mistake people make all the time when they select co-founders is to pick somebody who's really good at technology domain expert everything else good but not determined and that's very bad in general the hiring framework that has worked for me in my life has been to go for values first aptitude second and specific skills third and I think this is particularly true for co-founders the problem is that most people for regular hiring and especially for co-founders reverse this you know I need an engineer with a background in node and who has worked on iphone apps

about the worst possible way to choose a co-founder and then you really want people who are you want founders who are humbled who are humble and not entitled companies end up being reflections of their founders the culture of the the founders set is what will the company's going to look like for a long long time and I think a very common mistake that people make is to assume that well you know this founder I saw in this movie that is probably not even that based on reality was really arrogant so that's how to be successful

so that's how I'm going to come across it turns out everybody poisonous to the culture no once were quickly really bad and I have seen the reason I added to this slide is in the last year I have seen people who are kind of like more reserved like a little bit more introverted you know like closer to my side of the sort of a space kale than somebody else say you know what I need is like a really outgoing kind of arrogant co-founder because that's what success looks like and really that's just going to hurt

you and it has become this meme of this thing that people are trying to select for so I mention it okay so product this is like the most important yes everything that I said before except the determination and the self belief has to be even higher the way it normally works when you have two founders is or three founders is there'll be days where like one person is kind of down and kind of wants to give up but the other person is like feeling really good and saying we're going to do this don't worry about

it you pull the other person up and then the next day it reverses but the problem is if you're by yourself and you're going through the sine-wave when you're in the downs you could just end up giving up there's nobody to pull you up so we really look for people that have like incredible internal you know conviction about what they're doing it's hard though it's really hard I don't think I could do it I wouldn't try I don't think I could do it though I do yeah certainly for the first hires I mean later it

may get to the point like someday you will get so big and so successful that you just need this incredibly esoteric skill and then maybe you can compromise a little bit but only a little bit the first few hires like even the first ten hires I think you want to think of those people as like almost at the co-founder bar you really want people that are going to believe in the mission that share your values and that are just super talented super high out the to do whatever they do I I have almost never gone

wrong in my career hiring people who are inexperienced at a specific domain but super committed and super talented everything else they've done and trusting that you know maybe they'll be a little bit less productive for the first month but they'll come to speed quickly you know the leaders that I've hired for the different parts of my organization I've taken a bet on most of them that when they haven't had experience and it's worked out but it's that's the order I go in yes well one thing that it's kind of frustrating to some people I'm sure

about the way I hire is I don't I haven't I think there are people who are really good at interviews and I've gotten a lot better throughout the YC process but I am NOT one of these people who can say I want to work with you an hour every day for the next 10 years based off of two one-hour meetings and I know there are people who do that or claim they do that and good at it but that's just not me so I really spend a lot of time like if I'm going to be

hiring a direct report I'm really go spend a lot of time with them like we're going to go out for meals four or five times we're going to like you know spend many hours sitting together in the office we're really going to talk about like their life story in my life story what they want to get done in their life what I want to get done kind of like my philosophy behind Y Combinator and what I want it to be and why it matters and what they want to accomplish I hire really slowly and I

try to do it gradually as well it's like if you're going to be a partner at YC we develop this process where it's basically a year-long you know first you come in and you do office hours with some companies and then if those go well companies like you if a number of partners that are sat in with you like you then we're going to make you a part time partner for six or twelve months where you're kind of like like a partner but you do it much less of your time and so we get but

you're kind of like in our partner meetings and you kind of hang out and then if that goes well we're going to sort of like we have this sort of six month trial partner period we don't screw up that much once we actually make you a partner but we have a long time to get there and and we're pretty ruthless about you know if at some point it's just not feeling good even if it's someone we become friendly with which is why it's hard we just say we love you we want to help you find

something else to do but probably not here so that's the thing that's works for me there are people who kind of make these one hours snap decisions we do it for YC companies we decide very quickly but most of them don't work out anyway and you know when I'm with them every day so it's different yes all the time yes this happened for sure as well but you know most of the partners at Y Combinator have not been investors before this job and I'll give you here's this specific example a couple of years ago we

decide that we're going to start a growth fund although I'm an investor I am basically financially illiterate I you know like there are these people who can like build a model like you know say oh we think the company's going to do this much in 2023 and thus be aware of why and you know like I can I don't know the shortcuts in Excel so and that's ok at the stage of investing I do but it means there's a growth investor I really am well out of my league however what I do know how to

do is sort of say like okay I can't build your precise model but I think this is a company that's going to be worth a lot in five or six years for these reasons and I don't exactly how much I'm exactly what price we should pay but let's take a hard look but we had never raised a growth fund we had you know we have been investing super early and now all of a sudden we're going to be investing you know 50 plus million dollars at a time at a late stage in YC companies and

we had to like go off and figure out how we find professional limited partners because we need to way more money than we've ever deployed before and everyone's advice was the thing you need is to hire a growth equity finance person everyone's advice even my own partners but certainly like everyone outside they're like there's no other option no one will give you money without this you need a seasoned investor with a big track record it's a different world you need someone that's going to wear like cufflinks and suits and the whole thing and I met

some of those people and a hundred percent of them failed the values test and honestly I would say hi percentage of them failed the the aptitude test as well they did have the experience but any of those people that I brought into the organization to lead this very important new group that we have was going to be a big compromise on values everyone I met thought about that thought about that I was like you know what we're not going to do this so the guy we ended up hiring for this I had never been an

investor before his life at any stage he had been a CFO he was the CFO of Twitter but he had absolutely no experience as a growth investor and it was this thing that everyone said you just can't do this is the one area where you got to have an experienced person but I like any of them anything any of them weren't going to have good values alignment with what we're about as an organization so we hired someone absolutely no experience what we're doing and he's worked out fantastically well build a great team and I was

able to sort of get in more junior people that specific domain experience but he understands very fundamentally what we do and why and has built a wonderful organization in service to founders in the same way that we do it at a later stage that I don't believe any of the traditional growth investors would have done when the very early stage of my career you know no one experienced or talked to me at all so at that point I had to hire an experienced people that was not a choice that was a default I think that

many people who have done startups of understood what that's like you know the thing that always made me feel a little bit better about that is I was also unqualified and I was doing okay I was also you know I had no expandable or to know anything about business and know that much of my engineering I was doing okay so I was always I'm always skeptical of founders who say like I have no experience myself but I will only hire people with 20 years of experience there's some mental disconnect there so yeah I I was

always willing to you know there have been tug and there have been times where really specific domain expertise about something is important you know like I wouldn't hire a non-lawyer and say okay you'd be general counsel but but in general I think taking a bet on people has paid off for me when you were saying about the process of how you select your partner at this stage when your yeah it's a great question it's definitely hard one of the best pieces of advice someone gave me when I asked that that same question I was starting

my career is just be realistic that you're not going to hire that many people and that you're not going to get as much done as you'd like hold the bar super high because you're only going to get to hire three or four people anyway your networks not that big it's like really hard to get people to believe your vision you might not have that much money and so that kind of mentally freed me because I had been feeling like I had to hire a bunch of people and immediately to be taken seriously and then when

someone said that's just not realistic that's not going to happen so find a few people realize it's going to take longer and get really good ones and it's better to like hire two or three great people than ten warm bodies that really freed me and I was like okay you know what like I'm not going to hit the hiring goal that the board told me to hit because I'd have to compromise to do that and I'm just going to get a few people and if they're really good people I mean maybe we get less done

this year than we want to but it means we get more done next year and you know like still we're always behind a hiring plan every year YC is always behind a hiring plan it's just hard it just takes a while it never gets that much easier and then for example the story of Facebook Facebook one of the four billion dollars look a boards role is advice and consent the CEO supposed to run the company and I think if the CEO is doing a bad job that's when the board is supposed to step in the

board can fire the CEO sometimes less and less but theoretically the board can fire the CEO but beyond that the CEO is supposed to make the decisions and the board is supposed to you know be a sounding board advise and consent and I don't so like do I think a board should fire a bad CEO yes and I know that's like a little bit heretical in Silicon Valley beyond that do I think the board needs to give the CEO very wide latitude to run the company yes and I think most boards do if you don't

have a board though that you want advice from if you don't have a board that you intrinsically respect and want to listen to that means you have a bad board and you should fix that but I and I've had good board members I've had bad board members I have board members now but we don't we've never had a board meeting the but I like them all but I think you really want you want people who are there to support you to help guide you but will respect it it's your company you're going to make the

decisions equally equally important the you do have a CEO who is going to make a lot of the day-to-day decisions but every good co-founder dynamic I have seen and by that I mean like we're all the co-founders actually stay at the company there's huge compromise and even if one person is the CEO which they are like everyone is kind of aware of what everybody else feels most strongly about and there's huge compromise because you know it's like the co-founders own the company together and they should be running it together to some degree and the whole

like the dynamic you see sometimes where the relationship between co-founders starts really good but then someone kind of gets power drunk on being CEO and their investors their boards telling them it's their show the other co-founders and then kind of compromise goes out the window the other co-founders almost always leave and in all of the good long-term relationships I think there's really good compromise sometimes it's really bad and sometimes it's really good I mean there are definitely times when a co-founder should leave there are definitely times when someone is just you know burned out not

enough and we had a co-founder leave and I think that was good for everybody there are also times where it would have been much better if that founding team could have stuck together and gotten long for a long period of time so I think that depends on the specifics um you know nothing lasts forever forever but if you can have like a really good co-funding team together for 10 or 20 years that's that's quite valuable all right I will keep going if people have more questions though I I would rather answer those go ahead one

of the things was uh that out of the three yeah watch Brian Chesky on YouTube Brian is unbelievable at this Brian is like every time I leave Airbnb I'm over there yesterday every time I leave Airbnb I end up thinking like wow I want to go work here it's like so inspiring and he's like talk so clearly about it's like so good at making Airbnb feel important you feel important it's like politician level of skill about that you know I would just like I think I think people like genuine people and so I think if

you can if you are not naturally super charismatic which definitely I am not and many other people aren't either I think you can go a really long way by just being genuine in your beliefs your intentions and also your willingness to help others I think if you just like go out of your way to help people even if there's not an immediate transactional value to you that that develops really good relationships over a long period of time but I think if so I guys many times help someone even though the kind of immediate thing I

was trying to do with them didn't work out and it like has been unbelievably good for me years later when I've gotten too invested a company in a company that has now gone on to be super successful or they joined YC as a partner or something else so I think a long-term focus on hey this person seemed really great and there's not an immediate thing but I'm going to do whatever I can to help them and trust that it'll somehow lead to some value creation years down the road is good I think genuinely being excited

about what you're working on and trying to get other people excited about that rather than like this fake whatever sales tactic stuff people do works really well I think learning to communicate clearly honestly is a huge part of it as well yes you know that that is the decision that founders get paid to make it's it is different in every specific case probably yes but it really depends on the specific thing Alex Schultz who works at Facebook on growth I don't this is probably isn't his original idea but he talks about this idea of the

magic moment the one thing that people use a product and they love it so much but that's why they tell their friends about it Airbnb talks about not how do you get to a three or four or five star experience but how do you get to Ilyn eleven star experience like what is the most comically ridiculous thing they can imagine that they can never do but maybe they can go back from that and get you to a six star experience and that's so great that you tell your friends about this incredible trip you had on

Airbnb and that's how they grow so yes I generally do think trying to figure out what is the one magic moment what is the one thing that's going to be so good about this people spontaneously tell their friends to buy it it's important but there are trade-offs and if you have this one ridiculous thing but it makes the rest of the product suck so much it's probably not worth it I do believe this though by the way that nothing but a great product will save you it's true the general class of mistake that startups that

they'll make is they focus on everything but this and they do so well at everything else they do so well it fails in marketing they do so well at hiring raising money but they ignore this most essential tasks and then they feel they can't figure out why they failed and they have these sometimes fairly off-base conspiracy theories because they just can't understand so you think there must be someone out there working against them some destroyer trying to make them fail and they can't understand it hey there was just this one thing you had to do

and you didn't do it a piece of advice that my partner Paul behide gives is that it is more important to have a small number of users that love you rather than a large number of users that like you a lot like you a little and I think this is one of the most important counterintuitive ideas and startups most people think that they need to get a ton of users on day one and then figure out how to increase their engagement and that actually does sound roughly right the other thing you can do of course

is get at a number of users that like you would love you and then figure out how to find more of those people and the latter is the one that works almost always and it's and that's a counterintuitive and an important point yes depends on the product you know if it's an enterprise product where people are going to be paying you hundreds or thousands of dollars a year ten would be enough if it's a consumer product probably needs to be in the thousands I think one of my competitive advantages as an investor has been had

a super long time horizon and most investors even though they say this don't and I think it's kind of like a human logical fallacy because these things just take a while anyway but it is hard to go into an investment saying even if things work perfectly I'm gonna be holding shares for fifteen or twenty years most people just don't like that and so I think I have kind of found a tribe of other investors that think the same way and we have been able to run the tables on these kind of companies so I think

the issue is most people just no matter what they say don't honestly have a long-term time horizon yes [Music] you know this has been a problem of mine I've had a hard time throughout my entire career establishing people who really want to disagree with me a lot so I don't feel qualified to answer this but you are right to ask the question it's definitely a problem let me know if you figure it out I think it generally is useful to over communicate on these things and generally say you know person one is going to be

responsible for things X Y & Z person two is responsible for things a B and C and then when we get to the big strategic disagreements of the entire business we're going to resolve in the following way so and we have a lot of rules like that in place at YC so there's like our class of things that any partner going to do by him or herself there's a class of things that you can do if you can convince any three people to go along with them there's a class of things you can do if

your domains with one partner responsible for software one for recruiting one for you know finance one for legal one for operations and then there are class of things that require the entire partnership to vote as a whole and those don't come up very often because there's 18 of us and really painful to make a decision with 18 CEOs sitting around a table so we don't do those often but if we're gonna do something like raise that new growth fund that I talked about earlier like that one requires the whole partnership to get together and so

we have set these frameworks and it lets us work I think we're probably the biggest venture investing partnership that I've ever heard of and I think without some rules like that it would just be uncomfortable point when you're in early-stage startup don't attack it better - what would you think better to advertise strongly that you are gonna get this top or more clear that there are III think it's always really dangerous to sort of say we could do these seventeen different things and we have no internal conviction which to do I think it is like

I subscribe to this strong ideas weakly held approach of startups like there's one idea of those that I believe will be the best if I get new data Emily change my mind and I think the other extremes the extremes are both wrong and that's what I'd recommend yes you know we wonder about that I think less than people think so we have tried a number of different configurations but right now the current one is the hard tech companies they go through I see just with the software companies and they always come in and saying I

can't get anything done in 90 days and they always leave saying I cannot believe how much I got done in 90 days I think too often people use hard tech as a crutch and as an excuse for why they don't have to move quickly and that's just wrong so we just sit down with them on the first day and say okay we understand you're not a pure software company but let's figure out what you can possibly get down in 90 days and that focuses people and it means they get something done and they almost always

will say that was the most productive 90 days of my life so I think as a general rule the hard tech company should look more like software companies and it's generally a bad excuse when they say they can't and you know we have just had result after result after result in very different fields of people that thought they couldn't do anything on a three-month time frame being shot at how much they could get done so also on the phone recently with the company called ginkgo Bioworks they had existed for seven years before YC than they

did YC and the founder said to me that if they had just done YC initially they could have compressed that initial seven years into one year and so I think you have to move fast no matter what business you're in like everyone kind of wishes they didn't have competitors that would be nice but if you're doing something interesting other really talented people are going to be trying to do the same thing and they can work smarter or faster or harder or all of those than you and you the way you beat them is to just

have super talented people working on a very fast iterative cycle if I think I'm always interested in exponential curves and I'll just mention two that I think are important this first one is the idea of how quickly your product gets better so if let's say on every iteration cycle your product gets 10% better on average if your iteration cycle is a week and your competitors is three months you know you're going to leave them in the dust very quickly and the reason that I think most hard tech startups fail is their iteration cycles three months

which is basically rounds to infinity and they lose and lengthen and die and the very best ones operate on the scale of software startups it's a little bit harder but they figure out a way to do it that's that's really important the other kind of exponential trend that is very related to that point and I know that this is always a little bit of a landmine to say but I'm just going to say it anyway I think you should work really hard in the beginning of your career one of the things that I know one

told me and I later realized to be true is that studying and working hard it pays off like compound interest and if you do that in your late teens in early in mid 20s you have many working years ahead for the benefits of that compound interest to pay off and if you wait and do that until you're in your 50s or 60s or 70s you just have less working years left for that to pay off and so to degree you're willing to like you're say okay someone in my life I'm going to work really hard

and focus on this thing I think it's good to do that as a beginning of your career and then you get the benefits of that over a long time let me just make sure I don't have anything else really important in here to say okay we can skip the rest of that go ahead sure yeah well look investors aren't totally wrong about this the momentum is really important momentum is self-sustaining and one way to have momentum is to be growing really fast another way to have momentum is just to have continual breakthroughs in the lab

but that's not most hard tech startups and I have met many and been involved with some of the best one in the last few years most of them just honestly don't move that fast and yes it takes more effort to identify a fast moving company if you can't just look at a graph of daily active users but it is possible and I have found the more time I spend visiting these companies in these labs I can identify pretty quickly if it's a fast moving company or not I can usually go to one weekly research meeting

and have a pretty good pretty good guess and investors are not wrong that is really important you really do need that and the question is how do you evaluate it if it's an area you're not familiar with you know I try to sit down with every hard tech startup that comes into I see at the beginning of our program and figure out what their goal for the end of the program should be and you know like there's a lot of people who are willing to engage in a genuine open brainstorm with me and come up

with something that they think is a little bit too aggressive but they're willing to try and there are some people who are just like no I can't make I can't the next milestone for the next three years there's no intermediate milestone that makes any sense that's just how long it takes you don't understand science okay they always fail the you know like at one particular company was a cruise as a self-driving car company we funded in the wind them in January of 2014 this was before self-driving cars were cool this is an example of the

kind of class of ideas that all investors say they don't like then three years later they're begging us to find more of them because it's the new investment thesis happens again and again in history of YC and we found this company investors all think we're crazy self-driving cars are decades away and Google's going to win this is dumb what are you doing like we just want more photo-sharing apps YC sucks and you know sat down with the standard founder had no expertise in building self-driving cars or cars of any sort there's the former co-founder of

justin TV he's really smart and you know really know how to build a good company and we sat down we talked about you know what he could get done in three months and he had this idea and I was like well I don't think that's going to help you raise money I don't think that's going to build real momentum how about the other idea I had this idea that he thought was impossible and may have been right about and we compromised on was that he would have a car by demo day they could drive itself

down the highway from San Francisco to Mountain View and most people would have said I can't do that he's like I'm gonna go get that done like okay I'll ride and if you do and you guys got by the did morning of demo day and had a car that drove itself autonomously was hacked up body a for with you know computers hanging out the back and but it drove itself 90 days later and he raised the money made incredible progress still it just like he they had unbelievable internal momentum every week their car was getting

better on the roads and he wanted to try to raise an a round later and investors just said like well you know we're just we don't really think self-driving cars are gonna be a thing now it's maybe late 2015 let's say and no one's investing and I kept trying to help me investors no one was doing it and I was like this is so dumb I understand they don't have this graph of you know daily active users but if you go to the if you go sit in their office this is so clearly different from

every other self-driving car company could raise money could raise money so finally I was like it I'm gonna write you a check so I made my biggest investment of the year in that company and then like seven months later GM acquires them giant acquisition every investor that passed within the next year with Sam like autonomies our new thing we really want wants you to be funding more self-driving car companies this is the future but but had people been willing to just do the work and look at how quickly that company was progressing even though it

wasn't obvious externally from like a daily active users chart they would have invested and the problem is that most hard tech companies aren't in fact progressing quickly alright should we do the pitch practice now all right thank you all thanks very much Sam now I was watching from here and I noticed that all of you were paying attention and taking notes it never happens in my class so so Sam can I hire you to be faculty member of Waterloo engineering thank you it was wonderful and for you for paying attention there will be a quiz

later okay there is more to come Sam it bracele offered to show us how he would respond to a venture capitalist venture an advisor to a real startup pitch and then provide feedback now there is no script there was no rehearsal so it is the real thing the chosen startup is a lucid and the co-founder is four note chasms a day a Waterloo and during alum and he's going to tell us or rather Sam about his company and to pitch why Sam should invest so without further ado let's get per note up here and see

what happens [Applause] you are you going to vote this but I can't or you're happy okay you can join together hey actually me sit there so I can look at this life hi hi okay so why don't you just go through this and I'll probably interrupt with a lot of questions sure all right and by the way this is really hard to do and takes a lot of bravery so thank you for doing this [Music] so as lucid labs were met tech company we're trying to bring specialized care to underserved communities and potentially people who

can have doctors actually examine them basically the marketizing healthcare we're looking at dermatology so skin cancer is one of the most diagnosed forms of cancer worldwide in North America so more than 80% of all cancers diagnosed our skin and one in five people in North America will be diagnosed with skin cancer and so basically the way it works out is you have a lesion on your body is either benign or malignant you go see your dermatologist they give you diagnosis of benign or malignant and everything is fine if they're diagnosed correctly but there's still all

lots of issues with all the overlapping parts so if the lesion is actually benign and diagnosed as malignant and then the owner biopsies every year there's 17 million unnecessary biopsies which equals 17 billion dollars cost if the lesion is malignant and they diagnosis malignant it's a health care cost of health care to treat it and worst case scenario the lesion is malignant and diagnosed benign there's 20,000 people a year who die in North America and obviously if in the diagnosis stage at this late stage than the people you notice there's no point there's no return

and the problem is this they do it with visual examination skin cancer just like any other cancer that has stages of progression and it only becomes really visible at you know stage 2 or 3 here we call it negative stage 1 and then we term the other ones as 0 negative 1 and but biopsy is the clear way of diagnosis the gold standard so there's a knee-jerk reaction for physicians to just order it and remove their responsibility and put it on somewhere else what we do is we use computational imaging so we have sensing are

augmented with artificial intelligence where we can actually not necessarily replace doctors but we augment the doctors capability to be able to diagnose and provide care very high standards and we do them there's different things we do that but that's besides the point all just move on from here what we're really proud of is we can radially sequence cancer just like a genomic sequence where you can look at a sequence and say okay this clearly is this disease we can do the same with skin cancer through our computational imaging and an artificial intelligence and this is

a actually our product ADA which which is performing right now it's an imager and while it gives us that ABCD metric this is the state of the art in terminology asymmetry border irregularity color diameter so how big is it and this is balloon or not again done by eye and everything just based on the physicians recall of what they saw last time you were in if the D and then the most important part is e is evolution of the ABC DS and so our product goes through does those measurements quantitatively and suggest different cases that

are of similar similar diagnoses are another product we have this spectrum this is where the sequencing comes in right so this is basically work we're augmenting the dermatologist to the utmost level and would use we do some depth imaging so it's like x-ray without ionizing radiation so we can actually see into the skin and this is the first of its kind we're actually staging skin cancer as opposed to just saying if it's malignant or benign and this is the performance that we get today so on the left is what the dermatologists are capable of on

the right is what we do so true positive is how good are we are actually determining what is cancer as cancer true negative is how good are we at determining what is not cancer as not cancer and then the inverse of those so as you can see where the numbers are good the the sort of the competitive landscape works out like this if you plot it on the axis of artificial intelligence and diagnostic capability where you know away from a pack there there's a lot of companies who provide telemedicine teledermatology and there's a lot of

companies that just provide imaging and let the doctors make a decision and it's not really clear what they're looking at in making decisions on the North American market cap as well basically there are four hundred fifty thousand users potential users dermatologist nurse practitioners general practitioner family doctors we assumed five of these practitioners to a clinic and with a hardware and software as a service model we project the phone if it's a million dollar market cap in North America this market cap worldwide the one and a half billion dollars and that's just for skin cancer and

when you go to skin disease and other conditions it's about a fifty billion dollar when you go to cosmetic dermatology who's about one hundred billion dollar market this is the team the three people on top are the co-founders and you have a co-founding team we have expertise and artificial intelligence optics and photonics and ui/ux expert as well as know you must have the business and operations currently we're working on partnerships with Toronto Western Women's College Hospital and st. Michael's Hospital all in Toronto to run a basically trial at Pei trial with this device and so

that was great mmm way more went right than wrong there so I oh you could just give that presentation to investors now be fine but since we're here and there's a lot of people I'll make some suggestions it's come a long way it's really good I think the most important slide in the whole thing under the you know if you have a great product that's what you want to talk about is that you have a great product you have something that works so you have something that is better than human dermatologists in every metric so

the two things like this is great and then also early on I kind of got what you did but not until here so I think this would really benefit my number one piece of advice and number of use advice from most pitches is that you need investors or anyone you're pitching this comes down to the clear communication to understand what you're doing in like the first minute and it's always tempting because it's always tempting to say well we have this really big vision you know you eventually want to build the a eye doctor for everything

and you eventually want to help low income people have access to high-quality medical care by using computers and you can say that at the end of the presentation but you really just want to tell people very clearly upfront this is what we do and it's okay if this is what you do today and you're going to do a lot more later because what you do today is plenty revolutionary it's plenty big so I think the thing that would helped us the most is to add a slide here right after this one that says in one

simple to understand sentence you know something like we use AI to diagnose skin cancer better than the best human doctors like that's it just tell me that and now I've got a framework for you and what you do for the rest of the presentation and then I will everything else I will be thinking of it that way if I hear that's the beginning and that that is missing I would say from 95% of presentations and it's the thing that investors want the most it's just that early on so like this slide this is what the

first thing we start on you know I my mom is dermatologist I happen to know that skin cancer is working yeah it'd be great but like you know this is less important to me even if I didn't know it then knowing that you have built something that without a bio just purely visually with a computer can diagnose skin cancer better than the best human doctors that's what you should be talking about and slide to this is well this was a hard slide for me to follow I love the message which is that there's a huge

error rate but you could just tell me like there's a huge error rate it would be easier and 20,000 people die a year and we spend you know almost 20 billion dollars on unnecessary medical procedures and I would have gotten it not to be less confused this one again helpful not but so early on in the presentation I would rather be hearing about how the technology actually works how well it works where you are in the process of getting it deployed like I think one thing that was missing that was important is what is the

roadmap look like from here to you saving a lot of lives and saving a lot of unnecessary procedures that's all more important than this I think it's as important in a pitch what you leave out as what you say and this is valuable but it's just not as valuable as the burning questions that an investor will have so that again happen to know but even if I didn't I wouldn't be as interesting to me as you telling me okay so we have this thing here's the data that proves it works and here is basically a

road map of the next two years of what we have to do to get this in the clinic and to start making revenue and to start impacting people's lives this was this is good this is where you're actually showing the product this makes it seem way more real to people and you're actually doing something but I would have this before all of that I would just say now I'm going to show you what we built but before we do that we're better than human doctors today this is not a future thing this is not a

technology but this is working and then you can get into all the rest of this these slides I think are always terrible I have never seen one of these where the startup presenting is not in the upper right [Music] no-one's ever like worse on the two axes issues to show and and this one again like it's fine but I've never ever seen the slide where someone says we're a small market so I think what would be really helpful here is we have created this revolutionary technology it's going to have a huge impact and here is

how we're going to get it into the ninety thousand clinics that was for me the biggest thing missing there's other things I changed the emphasis on the order but the big outstanding question me at the end of this was okay like what what do the next two years look like what does winning look like what are you going to go do now what needs to happen and I think if you reorder things to talk about what you do clearly at the beginning that it works then tell me about the product everyone understands the need and

the problem and then spend the rest of the presentation with something that doesn't exist yet today which is how do you go from here to a multi-billion dollar company that impacts a lot of people and then the end then you can get into the big kind of war gonna be Dai doctor and revolutionize health care work just starting with with skin cancer that'd be fine that'd be great but again small comments super strong definitely I understood what you did and I was excited and I wanted to help and it sounded like a really good business

so that's that's all great thank you [Applause] okay I always try to end early but I'll do Q&A for a little bit longer thank you very much look the one thing I liked the one thing I liked when I was in school is when talks I ended early so I could go get work done so I will I will try to end this a little early but if anyone has a question we haven't gotten to yet I'd be happy to answer it yes you know my backpack of all the otherwise I pull it out and

show it to you I make a list every day of the things that I want to get done I every night before I go to bed for the next day and I have like three big ones at the top and then a long list of like 50 little ones like Hulst return and stuff and I have figured out what times a day I'm productive at different kind of tasks so like the morning is really important to me to get creative work done or to just get like high throughput work time I really have to focus

and I have been in that I have had the same productivity system for 15 years and it keeps working so I don't change it well I have like I write a plan at the beginning of every year for the whole year as well but like I don't look at that as much and I have this thing every day in my face of what I'm trying to get done and I look at that all the time I mean I kind of know like what the overarching priorities that I'm trying to get done are but I believe

you get there just like you're the momenta that the motto of a Blue Origin is something like step by step ferociously she's a great model for how you get things done in life and so like I have these big projects I'm trying to get done but I break them into very little pieces and every day I try and knock off a few of them there's no magic to this like there is no secret to productivity beyond like know what you want to get done relentlessly guard your time and get those things done and don't get

distracted on our stuff sure one big one is technology can make the world better but it doesn't that's not the default so you need people who believe the technology can make the world better and are willing to work to in fact make the world better another is that economic growth is central to a functioning democracy and startups are one way to generate that but if you don't have that you have is very zero-sum world you see this in my country right now where people are kind of angry because we have no growth and it's just

getting more and more unpleasant and I think if the thing that people in the private sector can do the most to help get the country back on track is to get economic growth back in the US we had kind of 200 years of unrivaled economic growth 100 years of territorial expansion we had a hundred years of new technology really working and people were mostly pretty happy and now we don't and because people are sensitive to relative differences in quality of life but not absolute differences if you don't feel your own life getting better every year

you are understandably unhappy or if you see your neighbors if you're on the wrong side of the wealth inequality split with your neighbors you're pretty unhappy and so I think fixing this I think that I think I really do believe that economic justice is going to be one of the most important problems of our time and I'm fixing that and all the different ways we can fix that that's a very central value to me you know having the u.s. continued to survive and be what I think is the greatest country in the world is of

great importance to me having the world get better every year having solving the environmental problems ending poverty these are all things I think YC can help with and that are really important to me that was totally unnecessary but thank you okay now should I open it Wow someone knows me well it's a hoodie thank you thank you all for coming to listen by the way this is been a lot of fun [Applause] actually there's no need I got every is basically the same same same talkers all my life there are socks which I never wear

but thank you very much I'm gonna wear these thank you so much [Applause]