Tuesday, October 4, 2011

Science: so easy a child can do it

A few days ago, a coworker and I were mourning the fact that huge chunks of our work could be done by a well-trained middle-schooler.

This is remarkable, because I've felt like this for years in an experimental lab. In fact, I started interning in an experimental lab at twelve, so apparently, microbiology is so easy a twelve-year-old can do it. And as I've seen more and more techniques I continue to discover that a kid could do it.

  • Making solutions - check
  • Western blotting - check
  • Flow cytometry - check
  • Cell culture - check
  • Pipetting - check
  • Minipreps - check
Now, I will say that kids (myself included) are not as good at troubleshooting, but given their lack of understanding of the process and terror of bigger people, they also make fewer careless mistakes. (Anecdata: I've had two summer students, a fourteen-year-old high-schooler and a thirty-something anesthesiologist. Guess who made more mistakes, because he thought he knew everything? Yeah. Not the highschooler.) And of course, all us grown-up geniuses are needed to actually plan research projects and interpret data meaningfully.

But I always assumed this was the way of the world in experimental work, and one day we'll all be replaced by robots anyway. In fact, our robot overlords are on their way, at conferences I've seen automated western blotting systems, pharma already has automated cell culture, automated patch clamp and high-throughput automated everything. What I did not know is that this is true of computational work as well!

I'm doing a three-month rotation in a molecular dynamics simulation lab, no wet bench work at all. But still, a big chunk of time is spent doing things like labeling components of text files, counting atoms and labeling/graphing bond distances in 3D graphics. My lamenting coworker spent three hours counting atoms in an attempt to match the number of atoms from one protein to the number of atoms in another. So even when everything is already run by computers, there's still low-skilled gruntwork to be done.

Once upon a time, I thought I'd always want to do benchwork, forever, for the joy of it. Now I've seen the light, so one day, I'll be a PI and only do the fun benchwork and spend the rest of the time thinking deep sciency thoughts. And I hope I don't come back in twenty years to describe my discovery that 50% of what PIs do could be done by a twelve-year-old too!


  1. I've had a mild crisis in the past few days because I started feeling like I'm doing a job that monkeys could be doing. All I do is pipette all day! So this post, on your part, is really well timed.

    It sucks that, after my initial learning curve the first few weeks here, where I learned how to do what needed doing for the protocol, grad school now feels like I've stopped learning anything at all and have just turned into a grunt.

    Do I really have to wait till I'm a PI till I feel like I'm doing un-monkey-y stuff? (That's the technical term for what PIs do, right?)

  2. Unless you're in the worst-lab-in-the-world, you shouldn't have to wait 'til you're a PI to do some more cerebral work! At some point, you will have to do stuff like analyze data and write a paper, or write a thesis proposal, or give a talk. These give you a chance to step back and synthesize - look at literature, chew over your data, plan some new stuff.

    If your PI isn't hands-on about getting you started on this stuff, ask her! Don't suffer the monkey work passively (but know that a certain amount of it is required, and sometimes you will go months with just monkey-work no matter what).