Publish or Vanish

A colleague asks what he should tell his students when they say:

What’s the point of formalizing all of this into such a long dense 10-page paper? why can’t we just put up our code and demos on github, maybe write a blog post on it, make some youtube demos and give talks, etc.?

Tell them they don’t want to end up like me!

Blog posts and demo videos don’t stick. I’ve been doing them for years and almost no one pays any attention. They don’t get cited by related work. It’s like they never happened. Writing a paper sucks, and you have to do a ton of bullshit to get it accepted, but it has two crucial benefits:

    1. You need a really deep dive to explain something new or complicated. Someone can closely study a paper and get 10x more information than any slide deck or video.
    2. Other researchers doing related work are obliged to read your paper if it is peer reviewed. Let’s face it, that is usually the only reason a paper gets read. Researchers are not obliged to cite non-peer-reviewed content, so it gets ignored.

Let’s see if I can take my own advice!

4 Replies to “Publish or Vanish”

  1. [Comment lost to failed captcha — sorry!]

    Hey, I have been working on a paper for months – it’s not an easy task for me – and I plan to cite your work. Not doing so would be tantamount to ungrateful intellectual dishonesty!

    I just confirmed that I cited you, four times, in my last paper (rejected) … I am starting to feel proud of myself for doing the right thing … then again (shame) I have not been proven worthy of publication. Consolation prize: one of my former students was.

    Renegade outliers’ pride, your point is well noted … and your colleague’s students should want even less to end up like me!

    Thank you,

    Enzo Alda
    Lakebolt Research

  2. – one video by DHH of Ruby on Rails started an avalanche.
    – a demo video at Strange Loop of Stephen Wolfram can blow your mind for weeks.
    – blog posts show up in Google searches and CodingHorror.com seems to be attracting a valuable audience over time.

    Signed,
    Structure and Interpretation of Computer Programs
    https://github.com/metaperl
    metaperl@gmail.com

Leave a Reply