Daniel Lemire's blog

, 12 min read

Being shallow is rational

15 thoughts on “Being shallow is rational”

  1. Jouni says:

    I would call such people “bureaucrats” instead of “hypergeneralists”, because the description sounds pretty much like what the vast majority of jobs for university graduates have always been. You are not hired as a bureaucrat because of your in-depth knowledge, but for your ability to fit in, to quickly learn enough to be useful, and to manage things efficiently. And often because of your network of connections.

    1. @Jouni

      Bureaucrats typically have very specific roles, and they go up the ladder within this role.

      The hypergeneralist I describe can one day write code, the next he could be recruiting a new colleague, and the day after that he could be meeting with a client. Then he could help setup a user satisfaction study. The day after that he might advise his boss on whether they should migrate to the cloud. He could have to comment on the new color scheme of the business logo.

      Even progressive companies like Google frown upon such things… but much less so than more conservative corporations.

      1. Jouni says:

        Maybe this is a difference between organizational cultures in our countries. Most of my friends who graduated from social sciences and humanities have generalist jobs. People from STEM fields, on the other hand, have mostly ended up at specialist professions.

        1. People from STEM fields, on the other hand, have mostly ended up at specialist professions.

          Did they? If you work anywhere close to information technology, then being a specialist for any length of time is a recipe for obsolescence in short order.

          1. Jouni says:

            If you work as a software engineer, you are already in a highly specialized profession. You develop software, because that’s what you were trained for. Your education doesn’t prepare you for designing bridges or working in a laboratory – those specializations require different education.

            The big difference between STEM fields and social sciences/humanities is the nature of knowledge. In STEM fields, knowledge is cumulative. The more you study something, the deeper you can go and the more options there are for you to continue studying that topic. And because it’s possible to go deeper, you have to go there and specialize before anyone considers you competent enough to be hired.

            Social sciences and humanities are more about ideas that complement and contradict each other than about cumulative knowledge. There are some jobs that require specialization in a particular field of study, but such jobs are rare outside the academia. Most people are hired because their education naturally prepares them for generalist professions. They are not experts in any particular topic, but in taking advantage of new ideas, tools, and viewpoints, even when the new ideas are incompatible with what they already know.

  2. Arshad says:

    An expert is a man who has made all the mistakes which can be made, in a narrow field. – Niels Bohr .

    I guess the world will be having very few experts, then, in the coming decades.

    1. We make mistakes a lot faster than we used to.

  3. jld says:

    That’s indeed a perfectly accurate description of the current zeitgeist, however it is not going to be beneficial for very long neither for individuals nor for society as a whole.
    This amounts to quickly “burn” every knowledge resource available while not providing for the renewal of such resources which requires much deeper and longer kind of work.
    It is thus an illusion to see this as progress or progressive, as a French speaker I guess you know the meaning of “feu de paille”.

    1. That’s indeed a perfectly accurate description of the current zeitgeist, however it is not going to be beneficial for very long neither for individuals nor for society as a whole. This amounts to quickly “burn” every knowledge resource available while not providing for the renewal of such resources which requires much deeper and longer kind of work.

      Don’t ever make the mistake [of thinking] that you can design something better than what you get from ruthless massively parallel trial-and-error with a feedback cycle. That’s giving your intelligence much too much credit. (Linus Torvalds)

      1. jld says:

        LOL, so you think the lead open source developers are “shallow”?

        1. Here is what Linus said at his last public appearance:

          I am not a visionary. (…) I’m looking at the ground, and I want to fix the pothole that’s right in front of me before I go in.

          Linus is not a philosopher.

          1. jld says:

            Either we are talking past each other or I completely misunderstood what you want to mean.
            When I say that Torvalds isn’t shallow I mean that, though he may not have “philosophical goals” and is focused on practical results, he masters a TREMENDOUS amount of knowledge which didn’t came to him by hopping around “miscellaneous tech tricks” but by DEEP WORK in the sense of Cal Newport’s Study Hacks:
            http://calnewport.com/blog/

            1. Your claim is that the current trend is burning through our knowledge capital and is not sustainable. My counter-point is what Linus said… we are going through an accelerated form of evolution, trying things out as far as we can. And that, as a whole, it is a lot better than “philosophy” because “philosophy” is based on an overestimate of our mental capacities. Human beings are pretty dumb and they progress mostly by trial and error. The more trials, the more errors, the better. Linus Torvalds is representative of this approach… he is a short-term guy who prefers to make a mistake today rather than think deeply about the matter and do nothing.

  4. Greg Linden says:

    One thing that comes up here is what employers value. At least in my experience, that’s mostly specialization.

    It’s true that this might change over the coming decades, but it’s also bumping against the tendency for people to discount that which they do not know. That’s a version of the Dunning–Kruger effect, people thinking things of which they are mostly ignorant are easy.

    I’m not disagreeing, especially not with your main point that willingness to learn new things and quickly adapt to new tools is critical to success. But I did want to point out that there is a little subtly here in what employers value.

    1. I think that many employers value experts in whatever came out less than two years ago… be it deep learning or Swift or whatever.