If someone wanted to give us that much money to invest, I'd probably do it, just to see what would happen. We could probably cook up some schemes for scaling. I wouldn't be surprised if it turned out to be a lot like scaling software. You run into a lot of unexpected problems, but you also end up finding ways to solve most of them. Plus you get unforeseen effects of scale, some of them good.
Please do not try to "cook up" small start ups. It's better people juice their own ingenuity on their own time, when they have something big at stake.
I grew up in a communist country and I remember how the state gave allowances to housewives in order to invigorate women's businesses. It didn't turn out that well; most women, like most men, are not interested in the ungodly effort and the labor of love that goes into enterprising.
[Edit: specifically, there came up a cottage industry of document fixers who could get you bankruptcy documents so you get to keep the federal investment AND get compensation. People looted the fund.]
Ask anyone who's ever submitted a YC application (and been rejected!) if they think that the selection process is in any way trivial.
YC invests in 40-45 of the best companies of thousands that apply. Even trying to imagine a selection process that chooses 40Bn worth of YC companies makes my brain hurt.
I think that no matter how hard you tried to stand it up, it would flatten out like warm jello to "anyone who writes a few coherent sentences on the application gets some cash".
for 40B, people will get into starting up for the salary.
don't start a company to make ends meet, start it to execute a haunting vision that you can't get rid of (ever wonder why people quit their jobs, often to their financial detriment, to startup? yeah, it's a curse.)
start it to execute a haunting vision that you can't get rid of
I agree with this - only partly, because it isn't necessarily in a person's control - but a haunting vision that you can't get rid of is a real asset. It's easier to devote insane effort to something that your intellect and/or soul are deeply invested in than to something arbitrary or trivial. I wonder whether luck is more of a factor in the latter kind of project.
In any case, I am having this experience these days, quite unexpectedly as I never set out to have it and used to feel dismissive when hearing people talk about "vision". I'm grateful for it. It makes it easier to crawl over broken glass.
I think a lot of you are missing the point. I could be wrong but I think the gist of the message is to look at what 40B into ycombinator would mean in terms of value added, employment created, etc. (with the wild assumption that it all scales). His point is to heavily criticize the 40B bailout of GM.
I'm as much for promoting startups as the rest of you, but is a massive, blind infusion of cash the answer? The thing about YCombinator, Sequoia Capital, Kleiner Perkins, DFJ, etc., is that there just aren't too many of them. There was an article on here just the other day heralding the end of VC as we know it, because the vast majority of the industry wasn't producing good returns.
Investing in startups is the most difficult capitalistic resource allocation problem you can get into because you have no "reputation" to go on. When you're investing in public stocks and bonds, you're usually dealing with companies that have an established track record, or an area of the economy where the variables are known. But that is by definition not the sort of thing we're interested in doing here.
The real question should be, how many startups can PG and team invest in and coach if money were not the limiting factor? Is money the limiting factor right now?
Ignoring the fact that YCombinator probably wouldn't know what to do with $40 billion (a lot of rice and beans?), the jobs created by such an investment would benefit about 1% of the country -- smart, risk-taking, well-educated young men who are willing to live in California.
The point of saving GM is only nominally to save GM. The point of saving GM is to save the industrial base of a big chunk of the US population, and hopefully keep places like Michigan from sliding into permanent and irreversible poverty. The benefit is greater than the sum of the jobs saved.
Well yeah this idea is silly, but his original point about saving GM (that it's also silly) probably still stands.
His figure for GM employees is old (2005). Last I saw, the current unionized hourly workforce is 60k employees (see link below). If you simply gave those employees $125k each, it would only have cost the government $7.5 billion. That kind of money should allow all those employees plenty of time to find other jobs or get training for a new skilled career. Then if you gave the retirement/healthcare VEBA what it needed, that would be another $20 billion, for a grand total of $27.5 billion... we've already spent more than that just keeping it afloat the last 8 months.
If they let GM die, of course some parts makers would have to scale down too, but the parts industry wouldn't collapse, because they still have the other car makers to sell to.
The argument about GM is the silliest part. I think it's somewhat plausible that YC could find a way to use $40Bn productively.
Again, the point of this bailout is not just to save 60,000 current GM jobs. The point is to save the GM, the GM supply chain, the local economies of several large industrial states, and the millions of people in an economic ecosystem that has built up around GM. Looking at only the 60,000 jobs at GM is taking an absurdly narrow view of the situation.
> the GM supply chain, the local economies of several large industrial states
I don't see how these are going to just disappear if GM is gone. There are other car companies these suppliers can sell to. Of course there will be disruption, and probably some job losses, but not the whole 2-million-person industry (to use the industry number). And if some of those jobs do disappear, maybe it's because demand for the product no longer supports them.
I know my perspective is different on this, because I'm 25, and honestly cannot fathom a lifetime with one company. My career will be a couple years here and there, and most likely I'll have more than one career as well. So it is alien to me that anyone wants to protect (or even prefers) this idea of having a single job, at a single employer, for life. Why is it so terrible that someone might have to find a new job or that a town might die out?
That said, I'm not unfeeling about this issue. I realize these are real lives here (my dad is a GM retiree). What I don't understand is why we are bailing out the company, and not the workers.
I'd rather we spent the money on retraining the employees, hopefully giving them skills in an industry that isn't slowly dying. And this generalizes to all those losing their jobs, no matter the industry. Stop trying to save the individual companies (bailouts) or whole industries (tariffs and subsidies) and educate people.
"I don't see how these are going to just disappear if GM is gone. There are other car companies these suppliers can sell to."
That's your opinion, but the facts are pretty clearly against you. Listen to the radio and read the newspaper -- the suppliers are already failing. And they're not failing because there's no demand for cars (GM sold about 9 million cars last year, incidentally), but because their primary customer has stopped paying the bills. That's what the government is trying to save -- not to mention the numerous American towns that have no economy without the local GM plant.
And before you argue that bailing out these towns is the wrong solution, try to remember that the world doesn't turn on a dime. Yes, a place like Youngstown, Ohio probably needs to find a way to survive without the GM plant. But they aren't going to diversify their economy overnight, and it isn't going to help them to let their primary employer die.
So maybe you're right, and some non-zero number of these suppliers would survive a total GM liquidation. I don't know. The point is, the bailout isn't designed to save GM, so much as it's designed to save the entire economy that surrounds GM. Everyone who tries to make the calculation in terms of a cost-per-GM-head is being simple, and is failing to consider the larger picture.
"And before you argue that bailing out these towns is the wrong solution, try to remember that the world doesn't turn on a dime. Yes, a place like Youngstown, Ohio probably needs to find a way to survive without the GM plant. But they aren't going to diversify their economy overnight, and it isn't going to help them to let their primary employer die."
How much can we afford to spend on people's romantic attraction to their home town? This country's had ghost towns appear when markets collapse and industries disappear before, it could stand to see that happen again. I happen to live in a college town where the second biggest employer (i.e. not the college) is hiring electronics assemblers every week, because it's in a growing market. Why can't a few people leave Michigan and move here?
It's probably wrong to think about this in terms of Youngstown, Ohio surviving the death of GM. People can leave Youngstown.
I think it is fair to see the 40B as simply saving the 60K or so jobs and little more. If a natural disaster took 60K lives, I do not think it will cause a prolonged national economic depression.
Likewise, I do not think adding 60K employed consumers will go a long way in fixing the current economic crisis.
Of course, saying that the 40B will "save" the jobs is in itself a big assumption. There is every chance that GM will run out of 40B and still shut down.
I'd prefer the US government revise the operating practices of agencies like the SBIR[1] to work more like the Grameen foundation[2] to help entrepreneurs of all kinds establish or expand businesses (not just the tech/media companies which are YC's focus) and then put a few billion dollars into it. Extending 40 billion dollars of microcredit loans would make a lot of sense to me in any kind of economy.
It's not users you compete for, but needs. And those are effectively infinite. The limiting factor would more likely be the rate at which technology can evolve. But it would be kind of cool if you ever reached that point.
I understand that the main point was to rightly criticize the GM bailout, but subsidizing YC would be wrong for essentially the same reason. Governments are unable to allocate capital as efficiently as markets can. This holds even if Paul Graham is the central planner, no matter how much you might hope otherwise.
(3) Build defensible advantages. Like, a castle with a moat to keep away all the crooks, scammers and favor-seekers. ("Make something people want to break into, but can't.")
Ha right, how the government allocates money, "Paul Graham" would spend $39,999,999,992.07 on coke and hookers and the founders would split a frappuccino.