An email exchange from fall 2006, posted with permission from Gui and Matt, who were kind enough to let me listen to their conversation. We were all strugglng students then, trying to find where our interests fit within the academic system at Olin that had taught us how to question everything, including it.

We still wrestle with tough questions like these as young adults, but I think we've also grown up considerably - Matt has been in two successful startups and is now involved with VC-fu in funding them, Gui is doing biomimetic robotics research that regularly makes the news, plus opening a hot new studio that's almost singlehandedly brought blues dance back to Boston, and I... well, I'd like to think I grew up as well.

Posted without editing, typos/capitalization and all.

Gui Cavalcanti:

so as i've been studying for the PDEs exam while listening to mark and ryan talk about high school AP tests, i've begun to realize something. i think i've really wanted to be "just a gearhead" for a long time. my training as a "mechanical engineer" was mostly done in a series of machine shops where the environment exuded "we don't need school to do stuff, we get along just fine", and i think a lot of it rubbed off on me as i looked up to the machinists around me.

as such, sometimes it's almost as if i want to do poorly in classes that don't "make stuff", and that i want to prove stuff can be done without math as often as possible. it's an... interesting point of view. i think it has contributed to feelings of fear and disdain that i've had for math classes. studying for PDEs has definitely changed that, though. i've now taken over 45 pages of notes in three days, and i'm understanding it all. i don't need to reject education for the sake of innate ability, they can coincide. nice.

Matt Ritter:

First off, I've definitely felt the desire to try and design entirely by instinct. I understand on an intellectual level why math and sciences are important, but it's REALLY tempting to just say that I'm good enough (or at least that I can be if I go do exactly what I want to do right now). It doesn't help that I've been unable to 'break through' on math like you have. (Tips on that would be awesome, BTW).

I think that a large part of the problem is that we've made the mature decision to take our learning into our own hands, but we don't really know what we're going to need out there (or maybe that's just me). I always hear that math and physics are much less important than they seem in school, but to actually make the decision to decrease their priority significantly is a big risk that I'd want some experience before making.

On the other hand, college is only a tiny part of my life as a whole. Actually, what percentage of your learning do you think comes out of college? More importantly, is it /possible/ to learn what you need to know on the job without a college foundation? Maybe our math, while not explicitly employed, is vital to undertstand concepts that will be important. The problem is that I have no idea.

Gui Cavalcanti:

on designing by instinct:

it's easy, and it's *almost* fulfilling. it's kind of like comfort food. you feel like you're getting things done, you have measureable output, life is good.

after you do it for long enough, though, everything starts to look the same. you get into ruts. you start wanting to do more, but you don't know how. you feel like you're building the same thing over and over again. you feel the need to challenge yourself, but challenging yourself with new projects just results in tweaking your existing "build-the-solution" process to fit, and you don't feel like you've learned anything. if you get good enough at tweaking in a certain build space (i.e., CNC mill-centric, lasercutter-centric, welded-tube centric, rapid prototyper-centric, etc. etc. etc.), it's entirely possible to build prototypes for just about any system you can think of.

a computer architecture project is building a computer out of k'nex. i build what should be a laid-up composite frame out of delrin and garolite sheets.

here's the thing, though. building approximations and prototypes always results in "good enough" standards. nothing's optimized, nothing's predictable before fabrication, everything hinges on the designer not forgetting what they think they know. would you fly in a plane you welded together? if you get to the point where you've built enough planes in a certain style to know you can fly safely in them, would you be willing to change build spaces and fly in the first one you build? how do you know their wings won't be ripped off in flight? do you have any way of knowing other than living or dying in the maiden flight? how do you know if a component is over-designed, and impeding the performance of your plane due its added mass?

math and physics makes predictions and optimizations possible. they let you build good stuff, instead of shoddy prototypes. right now, you pick random dimensions in solidworks that look right. once you get good enough, the dimensions figure themselves out - you just need to plug in the right ones.

on college: you learn everything in college. you'll probably never be in a place with such a high concentration of accessible diverse opinions and experiences again. the thing is, the *stuff* you're learning doesn't matter nearly as much as your *ability* to learn the stuff.

i tend to forget every specific of a math class two months after i stop doing problem sets, but i remember what particular methods are called and what they do for me. i can then look them up and relearn specifics, but the connection has been made in my head and i'm not afraid to learn more about them.

don't call anything useless, though. that 7/8"-20 tap sitting at the bottom of your drawer is fairly useless until you have to facemount a pneumatic piston to a plate, at which point it becomes the only thing in the world that matters. don't take tools out of your toolbox if you don't know what you're going to build.

design education encourages thinking in a certain way. drawing encourages thinking in a different way. solid modelling encourages thinking in yet a different way. each of these is equally valid, and the more viewpoints you can draw upon the more you can critique yourself and your abilities. math is an equally valid way to view the world, which will help you make certain cognitive leaps when attempting to understand a problem.