Comment by timr | Hacker News Reader

Comment by timr | Hacker News Reader

Parent Full thread

"This is a nice sentiment but as someone who has been a programmer for biology and medical research, it's not true. It's just like any other mediocre programming job, but your office is a folding table under a decommissioned fume hood."

It depends on your level of domain knowledge. I saw a lot of programmers start into the world of biology seeking "interesting jobs", but with no background in the field. What they don't (usually) realize is that if a research lab is hiring a coder without research experience, they have some non-specific automation job that needs done. Think: maintaining the website, DBA work, basic sysadmin tasks, etc.

If you have a thorough understanding of a research field and you know how to code, life can be better. There is the potential of great upside (and the problems are usually far more interesting than building yet another caching/database system for a website). But even so, you aren't going to jump ahead of the guy who has spent his life studying the subject matter just because you breeze in with the ability to write code. Writing code is the easy part.

That said, academia is always a shitty place to be if you only want to make money and you aren't a football coach.

I agree with this, but I guess I don't know who he thinks needs this advice. I assumed his advice was for the programmer lackey who wishes they had a more interesting job. I would assume a biophysics PhD in this day and age probably already has been exposed to programming, maybe even in Python. I mean, my sister has an MPH and she had to take SAS programming courses, and she graduated 8 years ago. It's 2011, not 1979... most people realized computers were pretty important decades ago.

Yeah, I'd generalize the advice to: don't be a programmer, be a domain expert who happens to write good code.

I think that's still extremely valuable advice in a world where "hackers" are obsessed with the whizziest language, version-control system, editor, database flavor, etc., and other sorts of largely irrelevant technical ephemera that ages like dead fish.

Meanwhile, nobody understands statistics. It's a problem.

I'd generalize the advice to: Math is really important, especially statistics!

I've found that makes a bigger bang to any domain expert then just programming ability.

> don't be a programmer, be a domain expert who happens to write good code.

A million times, yes. I work in RF design, which feels like an extremely stodgy part of electrical engineering. People think Excel spreadsheets are the bee's knees, even when changing one potential part in your circuit means re-entering its parameters by hand. Don't even ask them to sweep across different frequency points (which affects part performance).

I get a lot more benefit out of writing some short programs in python that allow me to quickly iterate through possible designs, and across a serious breadth of options, than these "Excel engineers" do out of their worksheets. It's frustrating to watch so many people ignore so much potential to improve their work and output.

If there is a one takeaway for me here, it'd also be "don't be programmer, be a domain expert who happens to write good code". Well said!

As for the frustration you have by observing people doing mundane tasks; I bet every person who has had some decent exposure to programming have felt this way at some point (i.e "re-entering parameters by hand in excel"). And this seems so prevalent that sometimes I wonder if people should be force-fed process automation in their daily tasks. Say, we start from schools where children are deliberately given such exercises where where they have to perform tedious tasks; and then they are taught to spot and come up with ways to automate in order to speed up. Organizations could also have such culture in place where they constantly look for things that may be unnecessarily consuming employees' times, etc.

PS, obviously I am not talking about programmers as it's in their mindset to try to optimize any work they're confronted with.

See also "Programmers Need To Learn Statistics Or I Will Kill Them All", by the same author.

http://zedshaw.com/essays/programmer_stats.html

I'd personally change that to "human beings need to learn statistics or I will kill them all".

"Humans need to learn statistics or _it_ will kill them all."

Bayesian statistics, right?

Not true in biology, at least. You're correct that most people realize that computers are important, but most biologists can barely use a mouse, frankly. Being able to code doesn't make you a great scientist, but it's one more tool in your arsenal that you can use to attack scientific problems, and it's a fairly rare one in many fields of biology.

Reply on news.ycombinator.com