I totally agree with Paul Graham. He’s definitely speaking from his own experience which turns out to be a very successful one. I have also noticed that employees even for the greatest company seem like gutless, castrated individuals compared with startup founders. I know as a fact that many highly gifted individuals but lacking the sense of adventure and guts that makes great leaders, hide themselves inside labs of the best companies.
Founding and running a startup is like trying LSD for the first time. It can be the most exciting and unforgettable experience in one’s life, or it can hurt and leave incurable wounds in one’s mind. So PG is like another LSD evangelist who had only positive experiences. However, early twenties can be traumatic in this regard, keep that in mind.
The attitude that only entrepreneurs have what it takes to do the job is insane. It is EXACTLY why so many people who start companies are lousy at running them in the long term, and have to make thier money by selling out and moving on. It is not easy to commit to the long term. In fact, “turning the ship” and re-making an existing software organization may be a much more daunting task than starting from scratch.
I do not know much about Y-Combinator, but it seems he is just trying to sell his product (or service, whatever). It’s like Joel with all his articles about FogBugz, he even said (many, many times) that one of the reasons that FogBugz is selling well is because his blog is so visited.
And you do not fool me with your attitude “Oh, I’m almost a full time blogger now, but I still build stuff”, you are sure going to use your blog in order to market your startup. And you should, you put a lot effort in it.
People still read Paul Graham?
As someone who has done both – run what would amount to a “micro startup” in today’s terms, and worked for major, very hierarchical companies, I tend to agree with Jeff. I think that Josh above got it right, about technology vs non-technology stuff in a startup. As a founder, you don’t have nearly as much time to do the cool technology because you’re dealing with mundane issues like business cards (and system admin, and…)
I think that several commentors are missing Jeff’s point, though. It’s not that corporate development is better than founding a startup, in direct opposition to Graham’s stance. It’s about the fact that neither mode is inherently “better” or “more natural.” So, corporate work doesn’t suit Graham, that’s his issue. It’s the fact that Graham is saying that you must either be a founder, or a loser. That’s a false dichotomy and I think that is what Jeff is objecting to.
Graham’s image of Big Software is skewed, too. While there are certainly many drones and companies that treat their employees like drones, the best companies have islands of enterpreneurship inside. They must do this. You can’t grow a company and a product portfolio by purchases alone. A viable company must develop new products and most have labs and groups dedicated to new product development. Then they need teams that can take prototypes and make them viable commercial products.
One of the proverbs we tend to use in our local Not An Employee culture is “Being incubated is fine as long as you like the smell of chicken ass.”
Or words to that effect. Graham is not talking about the same thing we are. He’s not even seeing it.
Thanks for calling him out.
Kinda agree. There is more than one way to do it.
And not all companies are the same, it depends on the company itself, and on the people behind it.
Making a go at it on your own can be just as risky, but even more rewarding in your 40’s and 50’s. The kids are grown, some of us are divorced and most of us know a HELL of a lot more about our business than a wet-behind-the-ears 20-something willing to go on a scavenger hunt. And most of us have learned better people skills than this self-important a-hole. He can speak for himself. In my worst years I was working for people like him - but still felt far from feeling like an “animal in a cage”.
I found the excerpt you chose interesting because it was exactly at the “lions in a zoo/on the plain” analogy that I bailed on Graham’s essay. Condescending and offensive even to someone who is a) not a programmer, and b) a business owner. Perhaps his next essay will be on how the best programmers have left programming to become venture capitalists.
I wrote a reply:
http://news.ycombinator.com/item?id=143433
If you’re going to quote me, I’d appreciate it if
you’d leave all the paragraph breaks in. The
paragraph breaks in an essay are as much a part of
it as the words.
“Paul Graham is the worst blogger I ever heard of.”
“But you HAVE heard of him…”
(Apologies to Pirates Of The Carribean)
Seriously, PG posts stuff like this to serve his own ends - it’s MARKETING.
How long will it take Jeff to start blogging about things that gently guide his readers towards a way of thinking that promotes HIS new web project?
It happened to Joel, it will happen to Jeff. Nothing wrong with it per se, as long as you realise you’re doing it.
Just read PG’s rebuttal - he hides behind a one line “get out” clause.
It’s a bit like saying “I don’t mean to be rude, but you’re fat and your mother’s ugly” - you’re still being rude.
Jeff 2 - PG 0
I find it interesting the way people read into things. Somewhere at heart, I’m much more of a skeptic. I read raganwald, PG, Coding Horror, Rands in Repose, or some MSFT-oriented blog where they fawn about the latest from the Redmond Firehose, and I ask myself about everything: “Is this true?” “Would this be useful?” I try to avoid reading anything personally, because nobody was writing solely to me.
I have also come to dislike ranting on my blog, after realizing that any post is potentially going to be a first impression. And I don’t want to unnecessarily alienate my future readers/friends by having that impression being drawn from some particularly angry or petulant post. Having a strong, personal voice is good, but so is kindness and empathy.
I think what I’m trying to say here is that I don’t like this post because it’s devoid of something to learn from, except for studying the way people interact and respond to things. “Paul’s latest essay sucks. Therefore, Paul sucks, and his future essays, now and forevermore, will suck because they’ll be about him sucking. BTW, Joel sucks too.” It’s almost like Steve Yegge’s old article on Perl which details at length how you should avoid it because Larry Wall sucks.
Is it making me a better architect / designer / programmer / sysadmin / DBA / businessman (pick any subset) to read this stuff, or is it wasting my time?
Give Paul Graham a break - it is called paulgraham.com after all - it is his right as an American to look down on people who are not entreprenuerial
Paul expressed a stereotype of companies based on their size. I get 10 messages in my spambox every day talking about how certain individuals aren’t good enough because of their size. That misconception can be softened by seeing more data that illustrates the wide bell curve of both size and entrepreneurialism of companies alive today.
There are a small minority of mid to large sized companies that consciously cultivate culture and work structures the let geeks to their thing AND manage to capture the business benefits of doing so - 3M is a classic example where the the RD folks get alot of leeway to build new ideas, and as a consequence 3M make a crapload of new patented things and keeps people interested in working for the mothership who would otherwise disperse to into individual startups. Many of the less business-oriented inventors would otherwise toil like captain ahab for years underfunded trying to launch or market one product instead of cycling through 10 of them.
One thing I hear from Paul’s essay is that there is a totally animal, primitive culture at work in his ideal startup. Everyone is coding, and no one is paying attention to building relationships within the company. Where is there trust that allows people to commit to ANY common vision and get through the inevitable conflicts of growing and monetizing and sharing the rewards of that monetizing?
With concious effort you CAN build at least a mid-sized company culture is something less bureaucratically structured than the death star and less chaotic than lord of the flies. If the social and economic system of that company is fair at 1-10 people, if you’re paying attention you can scale that system by adding new units or “cells” based as unified “mini-businesses” by sales territory, product lines, or whatnot. Each of these units are equipped to find and bring on new people, indoctrinate them into your mini society, and provide them resources and rapid exposure to new ideas from elsewhere in the empire that will cross breed like a thought-virus, making the end product stronger and differentiated from competitors.
Why don’t most startup companies scale without becoming Office Space Incarnate? They don’t have a vision beyond building a product - when they should be thinking about building a lasting COMPANY. Startup founders in the SV model feed off short-term angel or VC “junk food” money and vision - they build something brittle on someone else’s nickel, some small of them cash out to one of the corporate motherships Paul mentions, and they move on to their next startup. The remainder crash and burn. Building companies that last long enough to conceive and birth MORE THAN ONE product idea (YCombinator is one of these) is a lasting model that can outpace market change. A company like those YCombinator funds whose life support is based on one product risks losing everything overnight when competitors come along or market demand shifts - this is a basic truth about investment diversification. Paul is diversified because he’s harnessing the power of many entrepreneurs - he “owns” a small cut of many different founders’ efforts.
I wonder how Paul Graham staffs a startup with an attitude like that? Does he look at his employees as caged animals? Does he suggest to other founders that they should look at employees like that? Seems like an awful challenge to create a lasting startup if that’s the view of employees that you have.
People in glass houses…
Paul has it TOTALLY BACKWARDS about what you can learn as a programmer at startup vs. a top technology company.
You think you’re going to learn anything technologically interesting by starting your own company? Please. You’re going to spend a lot of time figuring out how to make a product that anyone will actually want to use, a lot of time doing all the mundane business/legal/logistical work of running a business, and a little bit of time here and there writing a Rails app. Unless your startup’s product is some highly technical niche, your company’s technology is going to be tragically unremarkable. You’re going to spend more time administering your wiki/version-control/webserver/SSL certs/etc than writing any interesting code.
If you learn anything technologically interesting at a startup that you hadn’t already played around with in college, you weren’t playing around enough in college. People don’t start companies to become technical geniuses, they start companies when they want to change the world or become a millionaire or something. You think Mark Zuckerberg is a better programmer than he was 4 years ago? You think he’s programming now?
Now let’s think about being a programmer at a company like, say, Google. All the boring sysadmin stuff is taken care of. There’s extremely good components you can use for your projects so you don’t have to reinvent the basics (RPC, storage, monitoring, etc) yet again. Your job is to solve big, hard problems and your toolbox is filled with the best of what the brilliant programmers around you have come up with. They’ve iterated many times and solved problems you wouldn’t have even imagined at the outset. And yet there’s always more to do, because the data gets bigger and the appetite for bigger problems grows.
If you want to take a big risk and maybe change the world or get rich, by all means start a company. But for someone like me who really just cares about the technology, I couldn’t be happier to be working at Google (for the record, I worked at a startup with some great, fun people, but it wasn’t for me).