Come on now, let's not go overboard here and compare YC to the Manhattan Project or Apollo missions. Doing a start-up is hard work, but we're talking about a slightly different scale here.
"Scale" is exactly the right word -- Apollo didn't happen because one person worked more insanely hard than anyone else had worked in the history of the universe. It was a large project with a lot of people (and a lot of science). Yes, a lot of them worked hard. But please, measure effort on a human scale. It doesn't matter whether you're starting a company or developing a new technology for space travel -- if you're pouring your whole life into it, the same psychological conditions can apply.
That is a rather silly criterion. All engineers build things that can kill people if they go wrong. That's what engineers do.
You think that social software can't be used to kill people? Do you know the number of people (mostly women) who are murdered each year by stalkers, most of whom are former lovers or spouses? Do you want to bet that none of those stalkers have used, say, Google to help locate their victims?
Do you think that having your identity stolen and your bank account drained can't lead directly to your death? What, do you live in a country where health insurance is guaranteed to all citizens? (Oh, wait -- perhaps you do. Never mind.)
"Toy companies"? Interesting choice of phrase. Here's what one minute's Googling found for me: A court case, Estate of Matthew C. Metzgar v. Playskool, Inc:
http://caselaw.lp.findlaw.com/cgi-bin/getcase.pl?court=3rd&#...
the plaintiffs' expert, E. Patrick McGuire, reported for the record that in one year studied, 1988, there were eleven deaths due to aspiration of small toys or toy parts by children... the plaintiffs submitted a CPSC estimate reported in the House Congressional Record that in each year from 1980-88, an average of 3,200 small children were treated in hospital emergency rooms for toy related ingestion and aspiration injuries. The CPSC also reported that between 1980 and 1991, 186 children choked on small toys, toy parts, and other children's products.
Frankly, from the viewpoint of danger and death, the space program's engineers are wimps: They do very dangerous things, but they only endanger a handful of people at a time. A toy engineer who specifies the wrong paint can kill hundreds of babies at once. Social software is pretty safe, but if the odds that a Facebook user will be killed because of a Facebook design error are as small as one per 100 million user-years [1], the company stands to lose 1 user per year, because they have 90 million active users.
[1] I have no data, but I doubt the odds are that small. I'd be astonished if Facebook hadn't already been a factor in thousands of suicides, for example. One of the reasons why humans are so incredibly talented at socializing is that it's fraught with peril.
Now you compare that to Facebook - if Facebook or Google goes down, who dies?
And this whole stalker point - wtf does that have to do with anything. Do you think there where no stalkers before the internet? The fact that people misuse something doesn't make it the engineers fault. If I designed kitchen knives and it is used to kill someone is it my fault? Hell no. But if I design a hatch door that doesn't open because of a design flaw, and people die, that is my fault.
That responsibility is the difference.
For some reason, this excellent comment reminds me of my visit to the sewers of Paris (they've got a museum down there!)
There's a quote from Napoleon in which he proclaimed that the redesign of the Parisian sewer system (which was commissioned during his reign) was one of the most important things he ever did. It's surprisingly hard to argue with that.
I think the test of the importance of your work is not how much people will be hurt if you screw up, but how much people's lives are improved if you succeed.
I notice you leave out the Manhattan Project. Through which nuclear power was discovered. And later those same researchers discovered Quantum Electrodynamics... which will likely be the next hardware solution for computer science.
Judged by the standard of how much it contributes to people's well-being, Google probably does more for the world than the Apollo missions did.
As for Apollo versus Google - come on. Search (PageRank) & improved web mail (GMail) versus safe commercial air travel, radio and satellite communication, safe structures, better roads, GPS, etc. Maybe there weren't advancements to Computer Science. But, there were major advancements in civil, construction, electrical and mechanical engineering.
The program spurred advances in many areas of technology peripheral to rocketry and manned spaceflight. These include major contributions in the fields of avionics, telecommunications, and computers. The program sparked interest in many fields of engineering, including pioneering work using statistical methods to study the reliability of complex systems made from component parts. The physical facilities and machines which were necessary components of the manned spaceflight program remain as landmarks of civil, mechanical, and electrical engineering. from http://en.wikipedia.org/wiki/Apollo_program
Separately the Apollo program had a cultural impact on the students and engineers of the time and even today. I know I am inspired by what the program did with such limited technology.
[EDIT: Some details and another quote]
I have no idea. This whole conversation seems like a big tangent, and perhaps I'm having a bit too much fun following it off into infinity. Shame on me.
Anyway:
Yes, Boeing has a more difficult task than many startup founders. Of course, there are also more engineers at Boeing, and they've been working on the problem for a much longer time.
Yes, it would be possible for a Google or Facebook design error to get someone killed. Leaks of personal or financial information are potentially deadly. Bad information can be deadly.
No, I can't give you a specific example of a design error (as opposed to a user error) that killed a Google or Facebook user. That's partly because Google and Facebook engineers -- like Boeing engineers, and most engineers worthy of the name -- take their work seriously and don't make many consequential mistakes. But mostly it's because our standards are looser, and reasonably so: The internet is a twelve-year-old invention, and when planes were twelve years old people crashed and burned all the time. They were experimental, and nobody expected anything different.
Standards will change. We're hearing stories of users who followed a Google map off a road and are now suing Google for damages. Right now those stories serve as hilarious examples of lawsuit-happy, stupid people, but perhaps in fifty years middle-class people will expect digital maps to be perfectly accurate, just as today they expect to have clean water and lead-free plumbing, and to always be able to reach an operator by dialing 911 on their cell phone.
Finally, I don't think this sentence is correct, in general:
The fact that people misuse something doesn't make it the engineers fault.
The problem is that there's no bright, shining line. Your kitchen knife example is far to one side of the line: Because kitchen knives are a very well-established technology, most knife accidents are caused by people misusing the product. On the other side of the line is, say, food that's been sweetened with ethylene glycol. That's pretty obviously a design error. And in between is a huge grey area, in which most software falls. Is Boeing responsible when a confused pilot pulls the wrong lever and fails to lower the landing wheels? [1] Is the guy who built the buggy online spreadsheet responsible when the civil engineer uses it to design a bridge that falls down? Well, that's why we have courts -- and why your startup needs to have a lawyer.
[1] I do know that planes used to have a little wheel-shaped lever to guard against this very problem. Maybe they still do. I forget where I learned that -- was it from Donald Norman?
When your brain decides how fast to make your heart pump, or how fast to breathe, or how tense to make your muscles -- it can't tell the difference between flying to the moon and building an ajax alarm clock. All it knows are what your emotional hot-spots are, the parts of you that matter to you... you are entirely capable of inflating the significance of whatever project you're working on just as large as the significance of Apollo to one of its engineers.
Likewise, it's just as easy for the Apollo engineer to say "I was just doing my job" as it is for me to say "Holy shit I'm responsible for my baby nephew's livelihood because his dad is a partner in this startup"...
Other people may not bat an eye when your web app fails, but you sure will.
My comment just pointed out that the definition of a cult used in the blog post falls down badly when applied to famous successful projects (which we can presumably agree were not cults), and is therefore either being misused or broken.