Do Things that Don't Scale

One more important thing to understand about Jobs "miracle" is that he was allowed to do what he was obsessed with. No idiotic "manager" or board member intervene with ideas of "cutting costs" on packaging or materials or any other fast-food technologies to reduce the product to the mix of cheapest and crapest ingredients imaginable. No idiots insisted that it is much more "effective" to just add a "theme" to and old code, to make it look like touch-capable (hi Symbian, WinCE) etc. No one said that re-implementing from scratch, because everything else is crap is costly and time-consuming, and so on.

There are lots of people with good taste, sense of style and quality (read Pirsig's book) and enough self-discipline to follow do what one is preaching, but it is an extreme rare situation when such people have enough influence to make things their own way.

I guess a half of YC business is about trying to spot an employ such obsessed individuals, no matter what exactly they want to accomplish. Early funding of even one of such guys worth waiting and funding hundreds of mediocre, especially if one has connections to channel or sell them off.)

Igor Sysoev was obsessed with efficiency (in that time c10k problem was actual), so src/core and src/os/unix should be taught in colleges (btw, the module system is already an over-engineered mess).

There are many other examples, but the big idea is simple - spot the right people, all that numerous micro-optimizations, while valuable, are superficial.

I had an odd way to follow this advice. I am learning to code and in 4 months I went from no knowledge at all to learning basic CSS/Html to basic C# to basic MVC to be able to launch and start selling a useful enough SaaS to acquire a few paying customers. I am just not a good enough developer to automate everything, I can't even code properly without Visual Studio helping me.

I am not dangerous, but I now know enough to be useful. I am trying to make my customers happy. The first feature I added after launch was something my first paying customer asked.

My product helps senior citizens to control their medications (what, when and how many pills they must take, which days, and alert when they need to buy more pills). I will start to put handouts on drugstores. Better than hoping elderly internet users know their way through google ads.

I spent 3 hours watching the first prospect to ever use my product taking a look at it for the first time (that would be my father). I left with literally 50 notes of what I should do immediately and some more thoughts of future features.

I am not a hacker, but I guess this helped me thinking more about what people want, because there was not much I could do from my own ideas and knowledge.

Speaking from personal experience, it's easier to pick up the traits of customer development/leadership/design with an engineering background. On the contrary, I've seen some of the smartest non engineer MBAs give up on programming. Not sure why.

I am not particularly smart, nor a MBA, or a McKinsey. I am learning to code to be useful, to build something useful, not build something great.

I would say that smart MBA try to learn to code to build something great, the next Dropbox, something that match their ambitions. When patio11 is your reference, not Bill Gates, or Zuckerberg, it is easier to keep your motivation. (Please, don't read this as me saying patio11 is not a great engineer, i couldnt know, just to illustrate the difference of ambition)

Most Americans with college degrees have heard of it, even if we never go near law school. Same for the GMAT and MCAT.