"They made a terrible life decision"

There's a myth that college degrees are unwelcome in silicon valley. But for every Mark, Sergey, or Steve that dropped out and founded their companies, there are thousands of engineers, analysts, designers, and programmers that didn't drop out and actually helped build those companies.

College is not a mistake, college is where you go to make mistakes. Including the mistake of choosing useless fields. Including spending six years of your life solving Schrödinger's equations instead of spending a couple weekends learning SQL. Yet there's a valid criticism that colleges are not adequately preparing students for the market. If someone wants to study Medieval Literature, there probably should be someone sitting them down and making sure they understand that nobody is hiring for that. Now my first thought is to blame the parents, but you can't count on that failure of an idea. In any case, isn't that why schools were invented? To be your parents while they're out working?

This problem is not exclusive to the non-STEM fields. Consider my field of study, Physics. There was never an explicit requirement to know how to code. Never mind that my field of study involved building protein simulations in Python. I know condensed matter physicists that had to learn Fortran because that's the language their advisor understands. Let me repeat that: the people whose job it was to help them make the best decisions were guiding them to learn the Latin of programming languages. They just gave up at trying to be actual advisors.

Okay, well, you're complaining about people not knowing how to code but giving all these examples of people teaching themselves how to code. Being self-reliant is a good thing.

Sure, but their self-reliance extends to accomplishing enough research to write a paper. That just requires knowing some For Loops and If Statements. Meanwhile, kids a decade younger than them are building data structures, learning version control, understanding necessary searching and sorting algorithms, and working on code bases in a team setting.

When deadlines are not as strict as producing results, you have less incentive to optimize your code. If the industry understands one thing, it's deadlines. The most important thing I learned in my first year at Quid was how to deliver quickly.

Yet time and time again, I've seen these same graduate students pad their resumé with these programming languages. I mean, they did use it in their research. As far as they know, it was good enough to get material results, so why wouldn't it be good enough for industry?

Because industry requires a higher standard of efficiency that academia. Unless you're hacking something together for a demo, you've got to write efficient, testable, readable code. Your experience is just not analogous.

Let's put our cards on the table, going into industry from a hard science is a pivot. You should be doing a post-doc now. Maybe you lost interest, maybe you weren't good enough. It doesn't matter. But the awful truth is that companies only care about what you can do for them now. They are not sentimental about your history.

You're in a better position than you think

It's harder to get through graduate school without knowing something about your field of study. At UC Davis, there were three filters a typical student had to pass through:

  1. A preliminary exam: A three-hour two-day test full of hard physics problems (and no notes/formulas/anything given)
  2. A qualifying exam: You present your thesis topic to a diverse team of professors for them to approve. Oh, and they can ask you any single question they want. (Feynman was famously asked to prove why the colors of the rainbow are in the order they are).
  3. A doctoral thesis, approved by three professors

There's also two years of classes in between, with their homework and tests to pass as well. And let's not forget the biggest filter of all: interest.

When you're an undergrad, at least half of your units are whatever random classes interest you. And the incentive to finish your program is strong when dropping out means all you're armed with is a high school diploma. In graduate school, if you drop out, you still have your bachelors, maybe even masters. That's not a bad place to be. Let's not forget, for all the talk of Larry and Sergei being dropouts, they were ABDs (All But Dissertation). These so-called dropouts had undergraduate degrees, a graduate degree, passed their exams, and were at the last step before their Ph.Ds.

In my experience, a graduate degree can almost guarantee you an interview. You may never touch your thesis topic again, but don't think it's not important. That stack of paper meant you saw something through. Something that took years. You got it done.

Companies need people that know how to do that. It takes an average of seven years between a company's first funding round and going public.

The drop outs are usually the founders or first employees, and they can get marginalized very quickly. The board knows they need smart people that have a high threshold for patience at solving hard problems. Startups are, for all intents and purposes, research teams. They're trying to figure out how to monetize an idea.

For certain jobs, having a doctor in the ranks can boost credibility. At Quid, I was an analyst in the client engagement team. We were all hustlers, trying to convince the heads of the world's biggest banks to make business decisions on things like betweenness centrality, for God's sake. With bullshit like that, it didn't hurt that half of us were Ph.Ds.

But you still have to nail the interview.

Remember, going into industry is a pivot. And that means you need to understand your competition, your weaknesses, and your strengths. You have a Ph.D, they can't meet you there. They have years of programming experience on you, you can meet them. You are smart, patient, and disciplined.

A couple times a year, we have a swarm of interns who join us at Facebook for three months. Many of them are from the University of Waterloo, who operate a co-operative educational program with various industrial partners. Instead of a typical four year programs, Waterloo students take five since a solid year is spent in internships.

I asked a colleague about this phenomenon, himself also a Waterloo-employee. He explained that being in the co-op doesn't guarantee you a position at Facebook or Dropbox. It's just that by the time they talk to these companies, they have dozens upon dozens of interviews under their belt. These kids understand the process. They've heard all the questions before. They know how to talk.

Start out with jobs you're not very interested in. Get used to the process. Listen to the questions and learn how to talk. Make mistakes until you get it. You will start getting offers, I guarantee you.

Then, when you're ready, hit the jobs you want.


comments powered by Disqus