Paul Graham Dropping Serious Wisdom

Every so often Paul Graham will email me something and say “can you read this before I post it?”. He did that last week. It was a talk he was going to deliver in Sam Altman‘s startup class. It was great. I told him I wouldn’t change a thing. I am not sure if he changed it before he delivered it, but what I do know is he posted it yesterday. And here it is.

I just went back to my emails with him and pulled these quotes out for all of you. These are some nuggets that I particularly liked.

On Investors – “our function is to tell founders things they will ignore”

On Selecting Investors, Co-Founders, Etc – “If you’re thinking about getting involved with someone– as a cofounder, an employee, an investor, or an acquirer– and you have misgivings about them, trust your gut. If someone seems slippery, or bogus, or a jerk, don’t ignore it.”

On Knowing About Business Before Doing A Startup – “The way to succeed in a startup is not to be an expert on startups, but to be an expert on your users.”

On Success – “Y Combinator has now funded several companies that can be called big successes, and in every single case the founders say the same thing. It never gets any easier. The nature of the problems change. You’re worrying about construction delays at your London office instead of the broken air conditioner in your studio apartment. But the total volume of worry never decreases; if anything it increases.”

On Finding The Next Big Thing – “If you think of technology as something that’s spreading like a sort of fractal stain, almost every point on the edge represents an interesting problem.”

Those are the quotes I called out in my emails back to Paul. They all resonate hugely with me. But the whole post is great. Give it a read.

#entrepreneurship#VC & Technology

Comments (Archived):

  1. Chimpwithcans

    Love the ‘On Success’ quote, and the stain analogy – Rings true. Sounds like a smart dude!

    1. fredwilson

      he isi always like to say “little kids, little problems, big kids, big problems” the same thing is true in companies!

      1. Chimpwithcans

        Sometimes it is so easy to put “living without a care in the world” up on a pedestal – take that out the equation from the start – truth is it is good to care, to worry, to work on fixing.

      2. kirklove

        Friends with older kids say that to us all the time. UGH 😉

        1. pointsnfigures

          they aren’t lying.

          1. awaldstein

            I can attest to the truth in this raising a teenage boy mostly on my own.

      3. JLM

        .Here is a hopeful thought. When they move out and get real jobs and they begin to be truly independent, you can watch caller ID.They get to a place where they want to impress you with their independence but still they want to use any vacation homes. They just stop inviting you.The empty nest is great.JLM.

  2. Barry Nolan

    The video is here. I think it’s even better. Replete with beautiful nuggets such as “Whenever you here somebody mention ‘Growth Hack’, mentally translate it to Bullshit’.” YC have ‘open-sourced’ everything they know about starting startups. Its excellent, and free, here: http://startupclass.samaltm

    1. bsoist

      I just happened to witness @barntude having a fit about growth hacking on Twitter on Saturday. It was priceless. I didn’t go read it, but he said he posted something on medium about the experience.

    2. Kirsten Lambertsen

      Thank you for sharing this! Listening is my favorite learning channel 🙂

      1. Richard

        Yep, read the transcript (and jot down a few memes that you took from it), then do the same for the video.Are they the same? They weren’t for me.

  3. David Semeria

    Top quote: …and you prosper only to the extent you doOr as I tweeted a while back: Riffing on Descartes: it’s used therefore it is

    1. awaldstein

      Great one.This is also so true:“The way to succeed in a startup is not to be an expert on startups, but to be an expert on your users.”

      1. David Semeria

        Total truth

        1. awaldstein

          Meeting with lots of startups lately vetting a bunch for a seed fund and this hits me every day.We can’t really know our model early but we damn well need to know the user and the market.

  4. kirklove


    1. fredwilson

      back to back this week. i’m programming for you!

    2. LE

      Get a room then.

  5. Bala

    Fantastic summary and curation thank Fred, I wrote a little about it because I am following the course just like the students and also hosting it in Reykjavik. What stood with me was the answers he had really a great resource for startup founders and those thinking and working in this world.

  6. Jorge M. Torres

    You have to read to the end to enjoy the best nugget:On Finding the Right Motivation — “At its best, starting a startup is merely an ulterior motive for curiosity. And you’ll do it best if you introduce the ulterior motive toward the end of the process.”The message to college students: Stay curious.

  7. pointsnfigures

    Ha, just went to the USV site, saw you posted that and read it. It’s full of great information. I am doing a college office hours at Ole Miss tomorrow, and will pull stuff from this. As I read it, I recalled that I had made many of the same mistakes in life myself. I especially liked: “starting a startup is where gaming the system stops working.” I hate it when I hear people say “Fake it till you make it.”

    1. fredwilson

      You can’t fake it in a startup for long. The emperor’s clothes and all that …..

      1. LE

        Otoh faking it does give you some runway to figure things out. Which is particularly helpful when you don’t know what you are doing and are going to make mistakes.

  8. p7r

    I love the idea of looking at the edge of the fractal to spot interesting problems. That said, there is still a lot inside “the stain” which is seriously broken too.

    1. fredwilson

      Yeah. It’s where I try to go. I think its easiest out on the edge.

    2. panterosa,

      The “fractal stain” is a fantastic image. Powerful, even more so for it’s odd word pairing!

  9. Rohan

    Man, he can write.

    1. fredwilson

      YuppppppppppHow’s B School going Rohan?

      1. pointsnfigures

        I need to go up and see Rohan at that Purple pavilion.

        1. Rohan

          Looking forward! 😀

      2. Rohan

        SO great, Fred. So happy to finally be here. Learning a lot, meeting great people and realizing that, with all the options presented to us every day, this is really a 2 year course in decision making and trade-offs.Still understanding the system but loving it!(On a side note, we had a lot of excitement here today as Prez Obama was down here and chose Kellogg as his audience. Sadly, I was among half the school that didn’t get the randomly generated tickets!My guess is that some major donor was an alumnus who asked him to do this.. :-D)

    2. JimHirshfield

      Write, he can.

      1. LE

        For the life of me I don’t know why he feels he needs to run everything he writes by a group of people for review:Thanks to Sam Altman, Paul Buchheit, John Collison, PatrickCollison, Jessica Livingston, Robert Morris, Geoff Ralston, and Fred Wilson for reading drafts of this.

        1. JimHirshfield

          Sanity check. Many writers have proof-readers and editors. Not everyune can right as perrfectly as you does.

          1. LE

            You mean “youse do” btw. Get it straight.…Except this is not “proof-readers and editors”. This is closer to a group of peers reviewing what you have written who have specific domain expertise and opinions of the subject. Fred is not an editor and Fred is not a proof reader. And if Fred wrote something and asked an office intern who was an english major to take a look or had a partner review something that’s to me quite a bit different than what Graham is doing here.Fred: Every so often Paul Graham will email me something and say “can you read this before I post it?”Besides Graham has the greatest forum (in some people’s mind) to get opinions on what he has written. He can just post the raw “essay” to HN and everyone and their uncle will fire away.I mean if you are going to claim that what you want to do is write and that that is your craft (enough so that you decided to step down from YC to do so) then why do you need to have others (other than, say proof reading if you want to do that) offer an opinion on everything you write (seems like this is status quo for all his essays, not just this one).This isn’t a book that is getting published. You aren’t Malcolm Gladwell. It’s just a blog post.

          2. JimHirshfield

            I think you’re thinking too much about this.So….scratch my “proof-reader and editor” and insert peer review just like science mags have.And…did he REALLY say he stepped aside at YC to be a writer (exclusively)? Even so, that doesn’t mean peer review is a sign of weakness.And publishing his work on HN for crowd-review kind of defeats the purpose, ’cause once it’s on HN, it’s fully published.

          3. LE

            I think you’re thinking too much about this.That’s what I do. I think to much. And it works for me. I make no apologies for that at all.…As for Graham’s future, stepping back from day-to-day duties at Y Combinator will enable him to focus once again on his writing. Before he started YC, Graham’s status as a thought leader was first cemented by his influential essays on topics such as technology and entrepreneurship.“I would like to go back to writing some more essays. When we first started YC, I could still write essays. But that was when it was just small. Then it got bigger and bigger,” he said. “It’s not so much the time, but it’s the attention. It’s the thing you think about in the shower in the morning. Now I will be able to write essays again, and maybe about topics other than startups.”Btw, note the fawning praise. This is exactly what happens in the press that makes people larger than life. Like he’s writing for the NY Times or WSJ etc. I hate that shit. I realize though that it’s not his fault that he didn’t ask for it. I just hate when writers say things like that and brand people.Even so, that doesn’t mean peer review is a sign of weakness.He’s publishing opinion pieces. Like when you see in the NY Times an opinion piece. He is not publishing peer reviewed scientific articles. The footnotes should not fool you.My point is simple. If you are confident in what you are saying then why don’t you just offer it up (once again this is not science but opinion) and let the chips fall where they do?I’m not claiming that if he needs to testify before congress or appear on a Sunday morning show it’s wrong to get some “help”. Or on your college application etc. I just don’t think this rises to that. Let’s hear what he thinks. Not what Paul with help from others think. Especially from a guy who graduated from Harvard of all places.

  10. Tom Labus

    So good about problems. They actually get more complex while sill carrying all the small ones too.

  11. Brandon G. Donnelly

    Love this one as well:“The best startups almost have to start as side projects, because great ideas tend to be such outliers that your conscious mind would reject them as ideas for companies.”

    1. Donna Brewington White

      oooh, good one

  12. William Mougayar

    I liked this part a lot, “You only need other people to give you advice that surprises you.”I think that’s where there’s a subtle difference between mentorship & advice. YC is a bit more like the masters-apprenticeship model, where the partners & other successful startups are the “masters”.Also, the part about listening to, or ignoring advice. You can learn by listening, by doing, or by making your own mistakes. It’s all good.

    1. Anne Libby

      There’s a lot to be said about the apprenticeship model.There’s plenty of things out there in the world of work you can’t learn in the classroom, only on the job.

      1. LE

        Part of the problem with young people (or any people) and the apprenticeship model is that it takes time. And most people want easy answers and things to be quick. Hence the reason they’d rather read and get the answer then figure it out on their own.I was at the dentist last year and he had to remove something. He had a few years before brought on a new dentist (to take over the practice) that was quite bright and even taught at Penn Dental School. (My dentist had bragged about this fact btw..) Anyway I normally don’t see “the new dentist” but this time I did. He couldn’t do what needed to be done despite all his education and even several years of actual practice. Old dentist (well not that old late 60’s) walks in takes one look and boom he’s done. Younger guy looks amazed old guy walks out of the room as in “been there, done that”.

        1. William Mougayar

          Exactly. So YC is a bit like a teaching hospital works. The interns & residents do the work, and the doctors watch & give feedback on their work.

    2. awaldstein

      more on this diff between mentorship and advice.i don’t think this is really a line that you can generalize.

    3. Frank Traylor

      I’m a mentor in a founder program. At the last meeting I typically ask “who heard advice from a mentor that they will remember because it was valuable” all hands go up. Then I ask “who heard advice that they will remember because the thought it made no sense” All hands typically go up.That is a great thing to learn; vetting advice. No one knows a company or its vision better than the founder. Great mentors know the limits not only of the founder’s knowledge but also their own. Advice from those mentors is golden but much advice is at best confusing and at worst counterproductive.How often do you find a truly great entrepreneur that doesn’t question common wisdom?As you mention though, the best advice lies outside common wisdom. It’s the advice that sounds brilliant, or crazy, or both. The kind of advice that you don’t pick up and act on but let percolate and expand your understanding of problems and solutions.

      1. William Mougayar

        Great comment. I consider advice as something you dispense only after you’ve got to know the company much better, and more deeply. Otherwise, as you said, it’s more in the realm of mentorship and ideas / education that percolate. The entrepreneur knows their company and product better than any outsider, but outsiders can point to certain things that may not be so obvious for the entrepreneur.

        1. Frank Traylor

          Completely agree. Entrepreneurs must have intimate knowledge of customers, company, product, competitors but everyone has blind spots that mentors can fill.Seems similar to board members. Good ones spend time learning about the company and focus on blind spots. Bad ones act as if they are compensated per word of advice.

          1. William Mougayar

            We agree.



        1. wot

          wot yas

  13. Anne Libby

    “You’re not sacrificing anything if you forgo starting a startup at 20, because you’re more likely to succeed if you wait.”This is wonderful.A lot of 20-somethings I talk with feel like failures because they haven’t “launched” something yet. This suffering is heartbreaking.Because in some corners of the world, “startups” have become the next Ivy League.

    1. fredwilson

      And YC is Harvard and Stanford combined

      1. Anne Libby

        Yes. And, there’s a lot of signal v. noise thinking we could all be doing about the world of “credentialing.”(And not just in startups/educational institutions. Ask a yoga teacher about credentials. Crazy.)

      2. LE

        And YC is Harvard and Stanford combinedOh wow I totally 100% disagree with you on that one. That’s only focusing on the upside not the downside. YC is not in any way shape or form “Stanford” or “Harvard”. [1] Not even close.Graduating Harvard or Stanford gives you a “plan b” in the event that whatever you do out of school doesn’t work. Whether it’s your first job or a “startup”. [2] [3][1] My guess is that in your case the first job you got in VC was greatly helped by the fact that you attended MIT and got a Wharton MBA. For much more than what you learned actually in class.[2] The majority of the world doesn’t know about or give a shit about YC. We know what it is and it’s got quite an excellent halo. But YC has only been around a relatively short period of time and not only that but Graham has also stepped down from running it to do writing. He’s not even at the helm, right? What does that say?[3] Not only that but as time goes on statistically with more bright people trying to be the lucky sperm it will be less probable to be the lucky sperm.

        1. falicon

          My argument would be that those that go through YC (and hence the startup roller coaster) actually end up with a ton of “plan B” options because they have been exposed to the gun fire (and likely learned a lot even in failure).Seeing what people who did YC (and failed) would be an interesting bit of data to explore…would you line them up against Stanford and Harvard grads (or dopouts?)…if you did, I suspect more of the YC group continue to “live the dream” and generate better than average odds at eventual ‘success’…

          1. ShanaC

            probably this is true of people taking VC from a number of specific funds.The problem is separating out the Stanford/Harvard from VC taking people without.

          2. falicon

            It’s always impossible to know the results from the path you don’t take…but I guess that’s what makes these debates fun 🙂

          3. LE

            I think that depends entirely on the person and the circumstances. For example if you go through YC and then you come up for air 4 or 7 years later what is the world going to be like then? What will YC mean (at that time) and will you then be settled in NJ and coaching sports teams in your spare time? And not be able to take advantage of the opportunity that you can get because of YC? (Not saying this couldn’t happen with Harvard btw but that “card punching” is awfully good and has proven so for many many years..)Harvard/Stanford and similar offer protection worldwide and way beyond “the startup scene”. A wide net of benefits.I suspect more of the YC group continue to “live the dream” and generate better than average odds at eventual ‘success’The business world is littered with things that are hot now and not later. YC has a short history. This is not to say that it won’t stand the test of time btw. But we don’t know.Here’s the thing (and I’m not a sports guy as you know). If you’re into sports and you get a chance to play for an NFL team or MLB, say for the Yankees (because Harvard is like the Yankees, right?) are you going to take your chance on the appealingly hot thing? As Clint would say “do you feel lucky punk, well do you?”. You know the reach of the major league teams and their halo even extends to me and my mother and neither of us gives a shit about sports.

          4. falicon

            Sports stuff == awesome :-)The reality of *right now* is that, if you are into building businesses and taking your fate into your own hands…YC (and the like) *are* teaching you more practical skills and if nothing else, you are truly learning by doing…Harvard and such will (probably) catch up, but right now, what you experience in YC is 10x (or more) useful towards building a real company/business (even if said business ultimately fails)…I guess that’s the point I’m really arguing here….but yes, back to your (real) point…sports and coaching is awesome! More people should just do that! =D

          5. LE

            That “awesome” is actually another bone I have to pick with this whole “scene”. You see to me business is and has always been “awesome”. For that matter programming (and I’m not a programmer like you are) is also “awesome”. I’m not good enough to earn a living from it but I can write things that have helped me make money. It just makes me feel good to write a simple script and make something easier for myself. Everything about business is awesome to me as well. I love thinking about it and dissecting it and would rather have 10 things and make $1000 from each than $10,000 from one deal at one time. It gives me a rush. Like when you connect with the ball. Would you rather hit ten home runs over 10 weeks or 10 in one game?So here’s my point. I got into business (and programming) let’s call it “organically”. Same way you probably got into sports. You were exposed to it nobody pushed it on you and nobody got you all hyped up about it as a cure for all of the evils. And it shows in your enthusiasm about it as well. But right now I suspect that there are many people that are involved in the startup scene that are there for the wrong reasons and aren’t fully committed to doing anything but being a star. Kind of like you are happy with sports and think it’s awesome but you don’t play on a major league team and you aren’t thinking anything is wrong with the fact that you aren’t either. And when they can’t operate a business or work in a business other than a funded internet business what will they do? You know most real businesses don’t have cool atmospheres like startups have. Because they are not being floated by investor money.

          6. falicon

            …and for the record, *YOU* are awesome too! 🙂

          7. Dave Pinsen

            Also,1) the signaling benefit of YC is even better than that of those universities, because YC is more selective and has less (or no?) affirmative action.2) as a friend of mine named Kevin once noted, founders who raise angel or VC money usually land on their feet, even if their startup fails.

      3. sigmaalgebra

        Okay, I’ve watched all of the first three YC lectures, read the corresponding Hacker News threads, and read the Paul Graham (“PG”) essay.Let’s see: P. Diaconis was at Harvard and later went to Stanford. Now both Harvard and Stanford are less good than YC? That is due to Diaconis? Wow! I like the work of Diaconis; he has some of the best work going on distribution-free statistics.Or, H. Royden was long at Stanford and wrote one of the best texts, clear, elegant, in measure theory and functional analysis.One of my favorite authors was long at Stanford, D. Luenberger. His Optimization by Vector Space Methods might be described as fun and profit from the Hahn-Banach theorem and is awash in nice applied math with lots of good applications, e.g., Kalman filtering at one time a big deal in US aerospace. He has a cute theorem that seems, finally, to make good sense out of the principle of least action, otherwise a bit obscure. in mathematical physics. He has one of the nicest, and nicely advanced and general, treatments of Lagrange multipliers, e.g., important in control theory, non-linear programming, integer programming, mathematical economics, etc. He has a good start on the Pontryagin maximum principle in control theory. Nice book.Harvard long had A. Gleason, a pillar in differential geometry crucial for general relativity. He also knocked off the Gleason bound that shows that no means of sorting positive integer n items by comparing pairs of items can run faster on average than something proportional to n ln(n). Since heap sort does run this fast, it is, except for a constant of proportionality, the fastest possible way to sort by comparing pairs.PG’s lecture emphasized *learning*. I can agree with that, and while I can like his essays, and so far have copies of at least 20 of them, I have to conclude that, for the goal of *learning*, Harvard and Stanford are a much better bet than YC!Indeed, PG indicated that he’d like to study physics. I can understand why he might! Well, for that he might very much want to read works by Royden and Luenberger at Stanford and L. Loomis, S. Sternberg, H. Whitney, and Gleason from Harvard.Indeed, PG’s lecture emphasized that what an entrepreneur in a startup really needs to know about is not *startups* but what the users/customers need. And the first two lectures emphasized the importance of building the *product* to meet this need.YC doesn’t really teach people how build a product to meet a need; some of the material at Harvard and Stanford can be much better at this. And for finding a suitable need, I was disappointed in PG’s The way to get startup ideas is not to think about having good startups ideas: you’ll get bad and plausible ideas that will end up wasting your time.To have a good idea, take a step back and don’t even try. I was hoping for more.One thing to learn about research is the importance of good problem selection; I have to suspect that this lesson also applies to the startups PG and the Stanford course are considering.Thus for advice on finding a suitable user need, I was hoping for more from PG.

        1. LE

          I was hoping for more.One thing to learn about research is the importance of good problem selection; I have to suspect that this lesson also applies to the startups PG and the Stanford course are considering. Thus for advice on finding a suitable user need, I was hoping for more fromI’m actually with him on this one.Ideas happen if you keep your eyes open. And you are able to spot an opportunity then use your skills to evaluate and decide if you want to take advantage of that opportunity (after you have vetted it).So it really does and should come from observation and inspiration.And there is really no way to predict when that will happen. To some people it happens all the time to others never.So people who are expecting some kind of academic plan of attack are going to be disappointment. I mean if you want to get into and graduate med school there is definitely plan that you can follow that makes it quite easy assuming academically you are qualified.But if you look at some of the great and even not so great businesses that have been formed they haven’t been done by someone thinking “what can I do” as much as “hey I’ve got an idea because I saw x and thought y”.Two examples (and I hate to use obvious suspects but you probably wouldn’t know the non obvious ones):Ray Kroc – sees McDonalds brothers are busy selling a boatload of his multi mixer and it gives him an idea that he is able to execute on.Thomas Stemberg (Staples) – saw office product flying off the shelf at the Makro Store in Langhorne PA (I used to go to that store and I remember that shelf)And so on. How many times have you heard “so I thought that…”How did you come up with the idea for Staples and what were the challenges of turning the idea into a business? MAKRO, one of the European pioneers of the wholesale club business, was recruiting me to become CEO of their US business. I did not like their approach, and suggested they consider focusing on office produts, which were flying off the shelves in their Langhorne PA store. They rejected the idea. My partner Leo Kahn and I ran with it!Here’s one on a smaller scale. Kirk Love’s former boss Morris Ballen (Discmakers bought cd baby) got the idea for when he reinvented Discmakers by seeing (in part) what I did in my printing business (he wanted to be my partner and I passed on that as I was in my 20’s and a few other he did it on his own.)…Morris realized there had to be an easier way to make and more affordable way to release records. In 1985, Disc Makers became the first pressing plant to offer complete vinyl packages to artists and record labels. This one stop concept – where the artist could buy design, printing, pressing, and packaging from one company – would end up revolutionizing the music market, and was the spark that created the do-it-yourself ethos that to this day permeates the indie music community. The launch of complete record pressing and cassette packages launched Disc Makers’ “modern era.”Morris was a really smart guy. Even a bit shady (which is an important quality btw.) And no, I didn’t “give him the idea” he merely thought (and I remember he said this to me) “I thought if a kid can do this with no experience (I didn’t have any) it can’t be that hard”. Not only that he used to come in to place his orders with us and asked questions and observed what was going on. He was able to put it all together in his brain and make it happen.So my point is he didn’t wake up in the morning and say “how can I pivot my business” he paid attention, observed, asked questions and then saw an opportunity and took advantage of it. By the way his business (which was started by his father) was debtor in possession when he gave me the first orders (was one of the things that scared me off I asked him “what does DIP mean on your checks” and he bullshitted me and gave me the wrong answer so when I found out I didn’t trust him..

          1. sigmaalgebra

            We’re not communicating: You are arguing against something I didn’t mean.Apparently we fully agree that problem selection is really important. Good.You gave a good explanation of this point with several good examples from business. Good.I mentioned that in academic research can learn that problem selection is really important. Same lesson, just different source of the lesson and examples.But, I never said or suggested and don’t believe that academic content, say, in the books of the academic authors I listed, explain how to do good problem selection, how to apply academic content to doing good problem selection, or even the importance of good problem selection. Right: Don’t learn this stuff in school, or at least not directly and maybe not more effectively than in business.But you and I seem fully to agree on the importance of good problem selection. Or, early in my marriage, as I mentioned to my wife, it’s possible to go really fast in, however, the (or a) wrong direction. Yup, it is. And I suggested to her that, really, picking a good direction was even more important — rare, difficult, challenging — than going really fast. E.g., I have a 9 pound kitty cat that can go really fast, but he doesn’t know one business idea from another.And you mentioned “vetting”. Good — now on problem selection you are nicely ahead of what PG did in his lecture.I have a guess (@JLM:disqus ): PG likes to have fun saying things that at first sound silly. One motivation can be, a reader thinks that, on the subject being considered, PG is being silly, dismissive, flippant and is trivializing. Then PG is somewhat hoping that his advice will get rejected and later someone will confess that in some sense his advice was good and should have been followed. Then PG can think “I told you so”.So, with my guess here, PG is in effect both tempting and daring people to reject his advice which actually can, even if buried, have some good content.Why? PG gets an effect of intimidation, that is, can just say stuff, just stuff, maybe basically with some truth but not clearly explained, and, from his track record of “I told you so”, have people straining to find the truth in his seeming trivializations. Maybe.Whatever the motivations, I’d prefer some plain speaking and clear explanations, maybe something like a statement, although there, with irony, deliberately misleading, of the Sidney Greenstreet character Kasper Gutman in the movie The Maltese Falcon.You are correct on “vetting”: Indeed, as we know well, Steve Blank goes on and on on how to “get out of the building” and test ideas, early on. At times, his advice is correct and powerful.I’ve had to do some problem selection:(1) In college I wanted to move ahead in math so took a reading course from the challenging, maybe even notoriously difficult, Kelley, General Topology. I gave a lecture a week to a prof. I started slowly, and the prof suggested maybe an easier book, but I continued. Then I did well.Ha! Later I took a graduate course which used that book, and in the first lecture the prof said that the homework assignment was just to work all the exercises!What a laugh! What a joke! Who the heck can even understand all of Kelley, much less work all the exercises! Since the class was just covering what I’d already lectured on in ugrad school, soon, to save time, I quit going to class. At the end, just before Christmas, I turned in a stack of solutions to the exercises but did not submit any of the solutions I’d done as a ugrad. After Christmas the prof said, “To be honest, I gave up on reading your solutions.”. He found no errors! So, picking Kelley as a text was a case of problem selection, that is, a challenge that was nicely hard but not too hard.(2) At FedEx, the company was held up because some Members of the Board of Directors were highly concerned that it would not be possible to develop good schedules for the fleet. As a case of problem selection, I accepted the problem, resigned from my position as the systems guy at Georgetown University but did finish teaching the courses I was teaching in computer science, joined FedEx, and designed and wrote the software, in a mad rush. The Board was pleased; COB F. Smith’s comment was that I’d solved the most important problem facing the start of the company. All along with the problem, I had to select what to include/exclude to get something good enough but in time — and did. It was important for my career then that my problem selection was good, a problem that was hard enough but not too hard.(3) In grad school again, I wanted another reading course. I picked a problem I’d seen in a course (right, sometimes can find problems just from daily experience) that was not solved there or, looking, in the library. So, I thought for a few evenings, saw a first-cut solution, and proposed the course, got my proposal accepted, and signed on. Then, after a handshake, I showed my first-cut solution, but it wasn’t very good. Could I get a better solution? I had no way to know except just some intuition that my problem selection was good enough. Well, two weeks later, from work in evenings sitting beside my wife on our bed while she watched TV, I found a nice, nicer than I’d expected, solution. I wrote up my work, turned it in, and was done. Two weeks for the reading course. It was fairly clear that, well beyond what the course requirements had been, I’d done something publishable that, really, in two weeks, had met the official requirement for a Ph.D. dissertation, but I used something else for that.Getting that solution was fairly important for my academic progress: First, the course credit was the last I needed for a Master’s. Second, that I’d done some good academic research polished my halo and, really, meant that the rest of my path through my Ph.D. would be fairly easy, which it was. So, via good problem selection, I’d found a problem that was hard but not too hard.(4) To support my wife and myself as we finished our Ph.D. degrees, I took a job in US military systems analysis. At one point the US Navy wanted an evaluation of the survivability of the US SSBN fleet under a special scenario of global nuclear war limited to sea. So, right, just do a simulation, of a large, simple random sample of all such wars and average! How the heck to do that? And, BTW, the Navy wanted their results in two weeks! Also, BTW, my wife had already gotten vacation reservations for us at a cabin in Shenandoah starting the day after the due date from the Navy. Should I take on this problem? Ah, a question in problem selection! Well, I took on the problem, did well, had my work pass technical review by a famous prof, the Navy got their results on time, and my wife got her vacation on time. There was a grocery store in Front Royal able to cut us two really nice Porterhouse steaks! I had the charcoal, etc., and we cooked the steaks on one of the picnic sites in the park. As we were finishing, a deer came up for a snack. Ah, maybe the park just had trained deer for that! In my solution to the problem, I had to pick what to include/exclude to get good enough results on time, and did.(5) My current problem is is from a case of problem selection; I have to hope that the user “need” is big enough and my work (solution, product) good enough.So, in these cases of problem selection, I used judgment, e.g., from experience in my craft, and not at all directly what was taught in courses.You gave some more examples, e.g., what Kroc saw selling milkshake mixers.But, once have selected a problem, then, for some problems, for the solution or product, what learned in academics can be from helpful up to crucial. Here, too, PG trivialized the connection between academic learning and having the tools to get a good solution, or, in the word of the first two lectures, product.Really, it appears that mostly PG wants to assume that, at least for work in information technology, the key, main, crucial core tools of the solution are just routine software. Well, PG, sorry, but in information technology, some much more powerful tools, yes, taught in some grad courses, are available.PG mentioned that in some cases technology is spreading like some growing fractal, and at the boundary of the growth one might find good problems. Uh, I’d say, that fractal boundary might be a way to find a new tool, e.g., the blockchain solution to the Byzantine generals problem, from technology to get a solution (“product”) for a good user need, e.g., Internet based funds transfer, but is not a very good way to find a good user need.Maybe most of us can agree that these days to find a user need for a startup, (1) do intend to exploit information technology in some since since it is still a great unexploited tool, say, something like the early days of domestic animals, wood working, metals, textiles, ocean sailing, iron, steel, steam, electric power, oil, internal combustion engines, radio, “plastics” (actually I never saw that movie!), etc. and (2) keep up on what people are doing and, thus, what new needs they may have.Or, get into the mix where the action is, and see what needs to be done that, yes, with vetting, is a good user need and will make a good startup idea. In a sense PG is correct: Hang out with other people, e.g., in college, as Zuckerburg did at Harvard, if can believe anything in the movie, and see what the needs are.But PG’s The way to get startup ideas is not to think about having good startups ideas: you’ll get bad and plausible ideas that will end up wasting your time.To have a good idea, take a step back and don’t even try. especially the last part, strikes me as minimal and even flippant, inviting rejection.How could it be that there have been successes such as Twitter where early on it was clear to at most only a few people, apparently would be bartender named Fred, knew just how big Twitter could be and why? And, why in PG’s experience have such a large fraction of the successes been from no careful attempt to find a suitable user need and startup idea?Well, there are a lot of people throwing darts. A few people try to aim carefully, but the number of people just throwing without careful aiming is many times larger. Net, of the darts that hit the center of the target, a shockingly large fraction were just thrown with little or no aim, view, insight, etc.But that observation does not mean that a person arriving at the dart board should not try to see and aim as carefully as they can. Essentially what PG is saying is a thigh slapper laugh at misuse of statistics.Or, that a lot of pregnancies are unplanned blessed events does not mean that a couple that wants a family should have unplanned pregnancies. Irony: PG’s family?

      4. Brent Halliburton

        One of the things you can’t discount but goes unmentioned when you do this comparison is the network. If you do YC, you are part of a community that can change your life. The best reason to go to Harvard/Stanford for grad school is not to attend the class, but to participate in the community. Future world leaders.

    2. awaldstein

      Talking with an Entrepreneur who is building an investment fund for rural Kenya based on rethinking economies from an entrepreneurial perspective.Crushingly bold thinking.

      1. Chimpwithcans

        I’m from Kenya – if you want any local perspective on the Kenyan economy let me know 🙂

      2. LE

        Crushingly bold thinking.Does “bold” equate with risky? How old is he/she, have they graduated from college (and which one) and most importantly what is their family situation to fall back on? In other words what is the downside of what they are doing not just the upside.

    3. LE

      A lot of 20-somethings I talk with feel like failures because they haven’t “launched” something yet. This suffering is heartbreaking.If the way you are relaying this is correct ie “suffering is heartbreaking” and not hyperbole I think those people are somewhat unstable lemmings who are simply following the crowd and actually deserve to be mowed over by people with better attitudes and more maturity. What they need is not comfort but perspective on life.You know what? It doesn’t get any easier as you get older.Same people that further down the road will feel like failures if they don’t have the nicest house on the street or the other guy has a bigger boat. This is something that you need to get a handle on early in life otherwise it will for sure bring you down. [1][1] I’ve seen this happen with marriages as well. Wife is married to a great guy, good father (in the sense of spending time with the kids) good looking all of that, decent living, but wife is unhappy because she focuses on a friends husband who gives his wife a bigger ring or more vacations or a new kitchen. On the other hand other people, more stable, are happy for what they have an don’t focus on what others have as a measure of their happiness.

      1. JLM

        .Comfort in one’s own skin and peace are not easy to achieve.The first problem is you have to recognize and acknowledge their existence.JLM.

        1. LE

          Comes from either parents or mentors. Essentially framing life’s issues and perspective. Super important.

          1. JLM

            .No argument with your observation, true.There is a subtle thing we may be exposed to when we learn how much is enough.How good do you want to be and are you playing against the market or are you wrestling with perfection and therefore wrestling with yourself?When I was building high rise office buildings — talk about an environment wherein you have to make timely and good decisions — I rapidly began to try to build the absolute “best” buildings I could within the constraints of the budget which was held hostage to the market revenue numbers.I would visit, catalog and measure the competition but ultimately I was the competition. I wanted to do it better than I had ever done before.I understand Howard Roark in Ayn Rand’s Atlas Shrugged because I got his desk somehow. It was a great run and I learned something about myself.JLM.

          2. Anne Libby

            “I would visit, catalog and measure the competition but ultimately I was the competition. I wanted to do it better than I had ever done before.”+1000. And it’s lovely to “see” you!

          3. JLM

            .How is the reading coming along?JLM.

          4. Anne Libby

            Ah, ha ha. Book group met last night. We’re going to meet again with a friend who’s a Latin American studies professor, for more cultural context…Thank you for asking!

        2. JamesHRH

          Because I recently had a humility-ectomy, I wrote a blog post that basically states that Paul Graham, The Bartender & Peter Thiel are all a little off base on this topic.

      2. Bruce Warila

        son, you will “be mowed over by people with better attitudes and more maturity”… somehow reminds me of the watch scene from Pulp Fiction.



    4. ShanaC

      it’s complicated.

      1. Anne Libby

        It’s complicated like body image is complicated. (Increasingly so for men, too.)There’s story being told out there like it’s one that everyone should aspire to.It’s not truth, it’s a story.

        1. JamesHRH

          You have your head up your ass if you aspire to what other peoole think you should do.Most of us do it at some point, though. I know I did.

          1. Anne Libby


    5. JamesHRH

      Everything that is awesome gets co-opted by people who seek success, rather than achievement. The Ivy League specializes in this behaviour.Michael Lewis put it best, although he was talking about Wall Street – http://www.businessinsider….





  14. Bruce Warila

    I can relate to Paul’s advice…- my one big success – I had significant domain expertise- my two moderate successes – I had moderate domain expertise- my three failures – inadequate domain expertise (I trusted my gut)

    1. fredwilson

      Well there you go

      1. LE

        Otoh anything that I have done I didn’t know jack squat about when I started to do it.One advantage that is clear from not knowing about something (and I think for sure you recognize this) is that you don’t have any bias or preconceived notions about how things are done. So you are more likely to experiment and take chances and not be held back by “the way it is done in this business”.For example I don’t know much about what you do in VC or how you have to coddle entrepreneurs. I get the clear sense that you are not the type of person to call up Daniel Ha and give him a third you know what because the photos don’t load as quickly as they should. (Hey wonder if disqus knows their users and customers?) Because you are a gentlemen and you want to preserve your reputation among entrepreneurs. Now me, otoh, not knowing how a VC is supposed to operate would be totally in his grill about the same issue (if I thought it was important). Unrelenting. Damn the torpedos.Now here’s the important thing. Doesn’t matter if I am 100% wrong. Point is I would try because I’m not you and I have no experience doing what you are doing. And if you get enough people like me trying some new thing someone will hit on something that actually works (given their particular expertise) that has been ignored by others.

        1. Bruce Warila

          Agreed… outsiders (that don’t know squat) are usually the disrupters. In each of my “failures”, I eventually acquired buckets of domain expertise. The next time I get into something that I have no business getting into (there’s always a next time), I am going to make sure I have a long enough ramp/rope.

  15. WA

    I have integrated the YouTube posted online lectures starting with Sam’s from two weeks ago in our Intro to Entrpreneurship evening class at the college this past week. Pin dropping 100% engagement.

  16. William Mougayar

    And the wisdom droppings continue in the comments of News YC by some who attended the lecture…

    1. LE

      And the wisdom droppings continueComments in YC tend to be lopsided though. Dissenting views or criticism that isn’t explicitly thought out and written with soft caveats gets downvoted and booed out. It can be like walking on eggshells.

      1. William Mougayar

        That’s ok though. That’s part of dissecting it. You can easily discard the bogus comments.

        1. LE

          But the comments aren’t bogus. They are just points of view that go against what the crowd (which changes depending on the comment thread and subject) thinks is the right thing to say.For example if the crowd (once again for a particular comment) seems to think that there couldn’t possibly be any reason for the government getting into individuals privacy then woe to the person that presents an opposing point of view. At least without cloaking it will a bunch of watered down adjectives. Or say you want to give your opinion on why Aaron Schwartz was wrong or even if you want to say “hey well he had some mental issues” etc. In other words certain types of speech (in our free speech world) are off limits on HN.Otoh the crowd tends to hang on the every word of known people (such as Andressen, Altman, PG to name just a few). As if nothing they say can be questioned and if it is you’d better be careful about what you say. I mean imagine if the rest of society was like that?To be clear: Not saying that the 3 mentioned aren’t questioned but they are definitely treated with kid gloves and not held to the same standards (to which I complain about) as others.

  17. Dave Weinberg

    My favorite, “”So this is the third counterintuitive thing to remember about startups: starting a startup is where gaming the system stops working. Gaming the system may continue to work if you go to work for a big company. Depending on how broken the company is, you can succeed by sucking up to the right people, giving the impression of productivity, and so on. [2] But that doesn’t work with startups. There is no boss to trick, only users, and all users care about is whether your product does what they want. Startups are as impersonal as physics. You have to make something people want, and you prosper only to the extent you do.”This is so true when hiring from the corporate world. This is the biggest adjustment – really in lifestyle/work ethic. Startups are on a different wavelength.Thanks for sharing Paul’s post :)Dave

    1. LE

      There is no boss to trick, only users, and all users care about is whether your product does what they want.That’s kind of like the “buck stops here” philosophy. When you work for someone else excuses can matter. When you work for yourself excuses don’t matter. In the end you are the one with the mess to clean up. Or have lost the account.What’s most amazing is how people that don’t own a business really have no seat of the pants feel for this at all. (I have in laws that are teachers and they are clueless to the point “he works for himself why can’t he…”)That’s why I go haywire when people tell Fred “it’s ok take off a month and stop blogging” and they clearly don’t understand nor do they have to suffer the downside of doing something like that.Which is why another quality that I think that is needed to be an entrepreneur is to be selfish. If you aren’t putting yourself and your company as the number one thing you are going to have a much harder time. (Other than health that is but above family because without money the family will suffer..)

  18. Eva Ongeri

    When I read this I immediately thought of the NYT article on the enterpreneurship marriage. It’s worth a read.

  19. falicon

    So many great nuggets…another source full of great nuggets and truths is Peter Thiel’s latest book “From Zero to One” (… ).At the end of the day though – this is the money quote (from your list) that everyone and anyone interested in starting a business *should* focus on and become obsessed with -> “The way to succeed in a startup is not to be an expert on startups, but to be an expert on your users.”It’s so easy to get sucked into the world, action, and idea of startups for the sake of startups…but all of that is really just an easy distraction…

  20. Leapy

    The greatest nugget for his audience at the time?”Good news: users don’t care what your GPA was. And I’ve never heard of investors caring either. Y Combinator certainly never asks what classes you took in college or what grades you got in them.”

  21. Nick Ambrose

    If you’re thinking about getting involved with someone– as a cofounder, an employee, an investor, or an acquirer– and you have misgivings about them, trust your gut….”This! 100 times over. Been there a couple of times and it sucks

    1. fredwilson

      me too

      1. Nick Ambrose

        I was going to write “At least it only cost me some of my time rather than much money” but then realized that’s actually worse than losing money in many ways !

        1. fredwilson

          way worse

          1. Pointsandfigures

            My blog post today was on that exact quote.

    2. Donna Brewington White

      I’m with you. On so many levels.

  22. FlavioGomes

    “How to start a startup is just a subset of a bigger problem you’re trying to solve: how to have a good life.” That was the biggest take away for me. With my startup years behind me and now in the growth phase; try to keep that perspective always in mind.

  23. PhilipSugar

    My favorite is that you can’t game the system. “playing house”.That one is just so true it hurts. I’ve never been able to put my finger on it. Its when you find that person that just has unbelievable credentials and work background for the best companies and they fail miserably. You think why?

    1. LE

      that just has unbelievable credentials and work background for the best companies and they fail miserably. You think why?I don’t think “why”. I know why.Let me explain with an example. I’ve thought about this for years has to do with the halo effect among other things. [1]Ron Johnson the architect of Apple’s store strategy later going to JC Penney and failing.To wit:…When Johnson announced his transformation vision in late January 2012, J. C. Penney’s stock rose 24 percent to $43.[10] Johnson’s actual execution, however, was described “one of the most aggressively unsuccessful tenures in retail history”. While his rebranding effort was ambitious, he was said to have “had no idea about allocating and conserving resources and core customers. He made promises neither his stores nor his cash flows would allow him to keep”. Similar to what he had done at Apple, Johnson did not consider a staged roll-out, instead he “immediately rejected everything existing customers believed about the chain and stuffed it in their faces” with the first major TV ad campaign under his watch. Johnson defended his strategy saying that “testing would have been impossible because the company needed quick results and that if he hadn’t taken a strong stance against discounting, he would not have been able to get new, stylish brands on board”Forget for a second whether the above is even true or not.Here’s the thing. Other than the obvious that JC Penney is not Apple the fact is RJ didn’t make decisions at Apple in a vacuum. Anymore than a President doesn’t make decisions without his kitchen cabinet. Consequently you are now taking Ron Johnson and putting him in a situation where you don’t really know how much of the brilliance of Apple Stores is because of what Ron Johnson did or what Steve Jobs did or what Tim Cook did or a million middle managers who might have an idea that they came up with that RJ blessed and agreed with. Unless you were a fly on the wall. Those people might not exist at JC Penny (and once again it’s a different business).So the question is who are you hiring? Are you hiring one person or a team and even so is the situation the even the same?The problem with hiring the halo is that you never know exactly who is responsible for what people have done and who has helped them.[1] Otoh we can be pretty sure that the Doctor that treats us is not asking for advice on what to do and holding meetings on your treatment (he is getting education of course) he is just using his own brain and making decisions and carrying them out. Nobody else is really involved at least the majority of the time. And the carpenter that does good work for you in restoring your house, well you see he has “good hands” which means the next renovation he will also have “good hands”. That’s the difference. People that build big and good things aren’t involved in all parts of the handywork.

  24. JLM

    .I am amazed at how little institutional memory exists about the methodology of startups. When I was starting businesses myself, I used to have a long checklist — before the Checklist Manifesto was written — and I used to follow it religiously. I used to write elaborate business plans. Many of these plans survived first contact with the market.When I got into the CEO coaching business, I refined my checklist and put it into the vernacular of startups. It is now about 140 questions long. I send it to startup CEOs, not as homework but as a means for them to organize their thoughts.Invariably they will say “whoa, Nelly I never thought about that.” These are brilliant persons who are doing business with the best VCs. This will be about some subject that I would have been able to identify from memory not some obscure double secret element.One client showed my checklist to a VC who promptly stole it. I didn’t really care but it was funny because we later met and he pretended he (his firm) had promulgated it. Not true but no big foul.In the “jockey, horse, course” analogy for business startups, I am flabbergasted how much wasted effort is expended in reinventing the wheel and how much credit is taken for inventing sex. [No, your generation did not invent sex or business.]If you want a copy of my checklist go to The Musings of the Big Red Car and look at “free stuff” or ping me and I will send you one.JLM.

    1. Donna Brewington White

      headed there now. thanks.

  25. JLM

    .On the subject of following your instincts as it relates to co-founders — perfect truth — there is also a corollary I call the “bridge” theorem.Not only do you have to feel comfortable, which may only be the dark side of the mirror of not feeling uncomfortable, you need to have something in common.Looking back on a third of a century of founding and operating businesses I have explored all of the relationships I have ever had particularly those that I would characterize as “good.” I find one common element.In every instance, we had some element of our lives in common. This common element became a bridge over which we could communicate and relate more intensely than in its absence.The bridge has to be some active element, not something passive like being “tall”.The best work relationship I ever had was with a guy who had gone to the Air Force Academy, a Zoomie. I went to VMI. Military school grads call their schools “trade schools” but not in front of others.We could relate to the common background but also the sparsity of words necessary to evidence a direction or command. I never had to issue an order because he didn’t need one. He could listen harder and better than anyone I ever worked with before or after.He could undertake big multi-million dollars rehabs of commercial and apartment properties with no more direction than a 5 minute chat. He developed a process that we could both relate to and which was used to document exactly where we were. The numbers were always perfect and the project was always under control. No surprises ever.In a big chunk of a billion dollars worth of projects, we never had a single misstep except for a terrible color selection I personally made. It happened only once. I had stepped into his area of responsibility and I never made that mistake again.The element of commonality does not have to be huge as long as it is real. One of the reasons I get along with Fred Wilson so well is that we’re both Army brats and engineers with MBAs. We may have different views of the same things but we can “relate” through that common background.Follow your instincts and find a bridge.JLM.

    1. Tom Labus

      Those relationships are rare and also unique. It’s tough to find people on the same wave length. Was he a protege or employee?

  26. David Miller

    The money quote for me was this about Mark Zuckerberg: Facebook is running him as much as he is running Facebook.This is spot on with my Founder experience. In fact, it got to the point where, among other things, I felt I had so little control over my life that I left day-to-day management; luckily as a co-founder, I still retained my Board seat, where I continue my work for the company.It turns out that Founders are indentured servants to their startups, and there is no telling how long the servitude will last. In my case, it was 11 years. The irony is that the more successful the startup, the longer and more difficult it becomes to extract yourself.No one thinks much about that when they get started. I know I didn’t.

    1. JLM

      .Stepping away from any company is always difficult and perhaps traumatic.Charles Degaulle said: “The cemeteries are filled with indispensable men.”We all need a life beyond work to make our work even more meaningful. When you ask a man who he is — he answers with a description of his work.JLM.

    2. fredwilson

      i did not either

  27. ShanaC

    Some of this is so smart, but some of this I find interesting – it is his companies growing some of these attitudes.

    1. fredwilson

      they aren’t “his companies”investors don’t control companies

      1. ShanaC

        Depends how active they are – In theory the CEO is an investor himself (albeit with sweat).

      2. vruz

        I think he sounds more like the founders are his kids. I feel like it’s not really an ownership thing, but a parenting, grandparenting or stewardship thing.

    2. vruz

      I think he sounds more like the founders are his kids. I feel like it’s not really an ownership thing, but a parenting, grandparenting or stewardship thing.

  28. Twain Twain

    Yes, I watched this and LOL’d a lot. Really looking forward to more of the Startup School series from YCombinator team and their connections.For me, his gold nuggets were: “Do what stretches you” and “Your objective in school is to satisfy your own curiosities and learn” because if you decide to do a startup later, that learning will be useful.

  29. Prokofy

    I actually dropped by to ask a question. And this post actually fits. Quora had this question the other day:…And I thought the guy’s answer was off the mark, but then I wondered what *is* the answer. And if I had to give one answer, I’d say “the rent’s too damn high.” It’s hard to live in NYC even if you have a bunch of room mates.You might disagree with the very premise — there are, after all, successful start-ups in NYC. Even if Twitter doesn’t turn a profit and doesn’t have a business model, so what, it’s life and air.I don’t think it has anything to do with people wanting to get paid or being unwilling to share academic research, that’s ridiculous, it’s shared all the time and PS they pay people in California. The first thing I thought of was Y-combinator. It’s not as if Y-Combinator expects to have interns working for free. They pass out cash.I liked the quote about knowing your users best. Really suits.

  30. fredwilson

    I think that’s a great point

  31. ShanaC

    Yes, but it also means you need to know a lot more other things.Fun thing about stains on fabric – you need to know not only about the chemistry of the stain – but also the fabric. Lets assume the stain is somewhat constant. The fabrics aren’t. And getting a stain out requiring knowing a lot about fabrics – which the tech side may not

  32. Donna Brewington White

    I think it means just that. And I’m thinking this will shape our appetites as consumers and as a society so that we will demand more customization and uniqueness from those who provide products and services, and we will even demand this of markets.