Thursday, 5 July 2018

Why is academia so inefficient?

So what is your benchmark that says it is inefficient?

Stuff takes time. Amazing amounts of time. I read somewhere that Google Plus took (something like) 500 programmers a year to code. When I look at Google Plus, it looks like a thing that takes my text and photos and sticks them on a wall. I feel like I could cobble that together in a couple of weekends, and I'm a crappy programmer.

Obviously, I couldn't, but to an outsider, it seems incredibly inefficient.
There's a statistic that IBM expected a programmer to produce 10 lines of debugged code per day. 10 lines !!! I could write that in 5 minutes. Why is it so inefficient?

The fifth Harry Potter book is less than (about) 250,000 words long. It took the author 3 years to write. That's 250 words per day. I can bang out 250 words between putting the coffee on in the morning and the coffee being ready. And it's not just JK Rowling. It's a very rare author who can bang out more than one novel per year.

How many programmers work at Quora?  10? 20? Let's say it's a few. But Quora seems very, very similar to what it looked like a year ago. So (10? 20?) programmers have worked for a year, and seem to have done almost nothing.

A few weeks ago I decided to paint my front door. How long can that take? Buy paint, buy a brush, put paint on.
So I went to Home Depot and bought some sandpaper, brushes, and some paint. I spent a weekend removing paint and making trips to Home Depot. I've been doing that for about the past 5 weekends. Last weekend I put filler in the cracks that are in the door, and bought some cement filler to fill in the cracks around the steps. If I'm lucky, I might put the first coat of paint on it on Sunday, but I think that's optimistic.  Let's say that's 3 hours a day for the last four weekends. I've spent 24 hours painting my front door (and for a couple of days I had friends to help), and I haven't even put any paint on it yet. That seems pretty inefficient.

Efficiency is output / input. But when JK Rowling is writing a Harry Potter book, or I'm painting my front door, or academia is doing whatever it's doing, it's not interested in efficiency, it's interested in quality.

Sure, if I were only worried about efficiency, I would have slapped some paint on the door. It would have taken about a 10th of the time, and it would have looked OK from a distance. Crap up close, but OK from a distance.  It would have been (say) 20% as functional. But taken 10% of the time, so my efficiency would have doubled. I'd have a crappy looking front door.  But it would have been efficiently painted.

I don't know much about the context of Crick or Feynman saying this, but they were both extremely intelligent. There were probably lots of things that they could have done in less time, but there was only one of each of them.

(The question has been edited, so a couple more thoughts. Are meetings necessary? Define necessary. All organizations have meetings. One unusual feature of academia is that its more democratic, and people want to get more involved. If you are prepared to let other people make the decisions, you don't need to go to the meetings. But you probably don't trust them, so you have to go and argue your case. You don't go to meetings to get informed. Crick, I would imagine, wanted everyone to do what he said, that doesn't work with professors. That's perhaps why he left.)


Thursday, 12 February 2015