A student's guide to startups (2006)
paulgraham.com
paulgraham.com
> So our official policy now is only to fund undergrads we can't talk out of it.
Speculating wildly on something we still haven't been able to figure out...
I was talking with an old colleague, and one of the biggest problems we've heard of in too many startups is founders with little-to-no experience who seem mind-bogglingly, clinically sure of themselves. Even in fields in which they have zero education or experience, even when all the highly-skilled specialists they hired are telling them the opposite. Even when they're told the good employees and customers will leave if they do a stupid thing that they have no business even considering. But why are so many of these people getting funded?
Going back to the article: imagine you're pitching, and startup experts who seem to be on your side are nevertheless trying to talk inexperienced you out of it. You can't make an argument that convinces them to stop trying to talk you out of it. Maybe they know something you don't? Nope, then you're rewarded for irrational confidence in yourself, and perhaps for manipulative fake-it-till-you-make-it behaviors.
Even if this wasn't the nature of the discouragement PG wrote about, given how influential PG's articles seemed early in tech startups, maybe a lot of investors since then have been selecting for, and encouraging, some traits that we've seen destroy more companies than they make?
If that's the case, was that part of some angel/VC formulas all along, to make some percentage of the investments be in narcissists, till you hit a high-functioning one? Or that, plus roll the dice on whether the necessary mitigating/complementary external factors align at random? Maybe that's not a bad strategy, especially when the investment environment is such that you can have an apparently skyrocketing property, and exit rich, before anyone realizes the rocket has deep-seated problems and is going to blow up.
Obviously, personalities like this have existed before tech startups, but maybe not as frequently enabled by professional investors?
Speculating wildly here, but the not-unusual clinically overconfident startup founder (to the point of self-destructive) is a phenomenon we haven't been able to figure out in any satisfying way, and only have guesses.
You're also giving way, WAY, WAAAY too much credit to "experts".
Two prominent counter-examples: AirBnB and SpaceX.
If AirBnB founders asked anyone involved in hotel industry, they would tell them it'll surely fail. The laws, the liability (what if someone trashes the place?), who will want to sleep in stranger's bed? who will allow strangers to sleep in their bed?
And yet AirBnB is amazingly successful.
Not because those were not valid concerns but because they were overblown.
Same with SpaceX: it was ridiculed on principle (what does internet guy know about rockets?) and it was ridiculed for their methodology (fail fast, iterate quickly, blow up rockets to learn from failures).
Armstrong, the guy who landed on the moon, testified before Congress telling them to not award SpaceX any contracts because they will surely fail.
Doesn't get more "expert" or more wrong than that.
Also, framing "fake it until you make it" as "manipulative" suggests that you're coming at this with some chip on the shoulder. The nature of learning is that first you don't know and then you know.
If only experts could do stuff then there would be no experts. Even Einstein didn't know physics.
I don't find it impossible at all. One example that everyone knows is Theranos. Elizabeth Holmes was a know-nothing with total conviction -- exactly the kind of young person Paul Graham lionizes in his post. She dropped out of Stanford at 19 to found a medical equipment company! When tons of experts believed that the samples Theranos relied upon were too small to yield accurate results! Which turned out to be true!!
There are, of course, a handful of businesses that have done quite well and defied all expectations in the process. But they are the exception rather than the rule. Most startups fail, and it's not at all obvious that being completely sure of yourself despite (because of?) your ignorance is an advantage.
It was basically a web 2.0 clone of VRBO which was already worth billions by the time Airbnb was getting started.
Why were people telling them that a proven business model wasn't going to work?
I wonder what this Sam Altman guy will be up to 17 years later
> Altman is the CEO of OpenAI
Got it.
Loopt mostly failed, and Altman's other founded company is WorldCoin, which also has basically failed. Altman's talent has been as an operator within the YCombinator rocketship. He hasn't shown he can build technology or great companies.
I mean, sure, you can make great profits when you put morality asides.
Which Gates did again and again.
People forget about this today because of the extensive PR of the last decades to wash the sins of the past, but Gates and MS were basically in the middle of a scandal every 6 months in the 90. Quite amazing since twitter did not exist to report it.
In fact, when IBM came to Microsoft to talk about licensing their programming languages for the PC, it was IBM who wanted a packaged deal i.e. get the OS from Microsoft as well.
And it was Gates who told IBM reps to go talk to Kildall.
Only when Kildall apparently blew them off Gates took a great risk by promising to deliver an OS as well. And he did it by buying an existing product to use as a base and evolving it.
It's all well document from both Microsoft people and IBM people.
And in a few years, there will be no documents stating anything wrong about Gates, only the praises.
¹ https://en.wikipedia.org/wiki/Stac_Electronics#Microsoft_law...
He’s been playing long game, working at this since 2015. That’s a long time to be playing without much of a payoff, but it’s all come to fruition now. That to me shows a lot of leadership ability.
7 or 8 years is only considered the "long game" in Nintendo or JavaScript framework time. I'd also argue jumping between substantially different industries, ideas and companies is NOT the long game, rather he's continued to play a game, and with significant systemic advantages found something that might be a moderate to large success.
(Yes, some founders obviously hit jackpot)
My point being that the sort of intrinsic motivation you are describing gets sort of overwhelmed by huge benefits packages and total comp.
Seems like this was every other startup in the 2020-2021 period. Zero experience founders raising a ton of money. I wonder if this is shifting a bit with the financial markets tightening up.
Yes, younger founders are low maintenance, but older founders usually have a large network to tap into for fundraising.
In the end, it's all situational. Younger founders are lower maintenance to the balance sheets and have more potential but it's more risk. I'm not a VC so I couldn't say what goes into the decision making there other than "he's younger and smarter so I'm betting on this horse".
I've noticed that an entire generation of essentially 0% interest rates has skewed this substantially. I'm in my 40's and I've scrimped and saved and have very few "premium lifestyle tastes", while many 20-somethings have lived well beyond their means financed for free their entire lives. Their expectation for the baseline standard is much higher than 15-20 years ago
There are a lot of people work for a big company for a while, or go into an industry, then go to a start up.
I don't have any data to show this is true, but I suspect that it is.
Not to mention that the world is profoundly less allowing of entrepreneurship than it used to be. I graduated from college in 1999 with about $25,000 in student loans with a $340/mo payment and my rent was $250. I was able to get by on a $10/hr seasonal furniture moving job. Today those numbers would be at least $50,000, $500/mo loan payment and $750 rent, so a person would need $20+/hr just to survive. Which means a real job and 40 hour workweeks with no time or motivation left over for a side business. Same city, double the population, half the independence. Trapped.
I'm disillusioned with it all because nobody ever solved the funding problem. Kickstarter is opinionated in what it will fund. Crypto turned into a Ponzi scheme instead of a way to fund startups. So it's probably over, and I'm just waiting now for AI to disrupt society so completely that it's forced to adopt UBI or face 50% unemployment.
First, this is one of those things which is easy to say but for many (most?) people hard to implement. Sure, some people's family may be able to give them the money they need to go to university but this is not the case for most people, and in this case it is really hard to earn enough money _before starting_ to pay for your university education.
The second is that debt can be good (our society is literally built on it) in many cases (e.g. home loans, provided you can afford to pay them off) if you are not in a highly asymmetrical power relationship with your lender.
> debt can be good (our society is literally built on it)
I would argue that our countries increased reliance on debt over the past 50 years has decimated the lower class and harmed the middle class while the wealthiest class has seen their influence and ownership percentages surge.
This isn't terrible advice for dumb new grads who don't know better, but there are very different types of debt. Consumer debt is almost always bad, but capital debt can be good, preferred or necessary. The challenge is something like a college education could be either.