How Pixar Hires
edutopia.org
edutopia.org
"The core of skill of innovators is error-recovery not failure-avoidance."
"Mastery in anything is a really good predictor of mastery in the thing you want done."
"The proof of a portfolio versus the promise of the resume."
"We want somebody who's more interested than they are interesting."
"People who are interested are much more willing to work on the communication as a destination not as a source -- and breadth, a broad range of experience in the world is the thing that fuels that."
"Depth, breadth, communication, collaboration; of these four collaboration is the most important"
"Collaboration for Pixar means amplification -- the amplification you get by connecting up a bunch of human beings who are listening to each other, interested in each other, bring separate depth to the problem, bring breadth that gives them interest in the entire solution, allows them to communicate on multiple different levels: verbally, in writing, in feeling, in acting, in pictures, and in all of those ways finding the most articulate way to get a high-fidelity notion across to a broad range of people so they can each pull on the right lever."
For a programming business which needs, let's say, a Java programmer, I think it matters more (to the recruiter) that the programmer knows its fu well, than being very interested; basically, the theory doesn't quite apply to existing businesses which use common technology.
But, as a programmer myself, I've only met a handful of other programmers whose knowledge I'd value over their interested-ness. I'll probably never have the opportunity to hire such a person, so when I interview I look for competence and learning potential (curiousity, humility, ...).
Also, I regard programming as a maximally innovative field. There's no reason in principle to do the same thing more than once in programming, and the practical reasons all seem to derive from a combination of mistakes and misfortune.
I love to mentor. At work, the fields I'm much more experienced in than coworkers I try to spread my knowledge but there's only so much I can force on them. And they seem reluctant to ask for help. They want to do it on their own. I notice them struggling for hours over something we could resolve in 15min working together. I want to boost them up, not make them look bad/one up them. I seem to be failing...
How can you make people into collaborators?
Unfortunately, unless your employer has similar values, it'll be challenging for you to create a similar environment.
Here's my beef against him: Lack of trust. If he says to me: "Do X", then expect me to do it.
If hours have passed and I still haven't done, and I didn't ask for help, then something is wrong with me. I have a task and I am not completing it. But you got to give me the chance to actually do it.
For example, he asked me to test a new mod_rewrite configuration on a server, and that I should use curl for it. I was going to do it anyway, so I kind let his comment go from on ear through the other. But then he said, hey, you must send an specific header because of the way our network is configured. That was actually useful, because I didn't know about this configuration.
So I typed "man curl" on the command line to know how to send a header. He said to me:
- No need to look on man, I will tell you what you need to send the header. Type -H...
I interrupted him and said:
- Wow, let me discover how to do this myself. Let me have a chance to learn.
So, here's the thing. There are struggles and there are struggles ;) If I am really struggling on something that is simple, give me guidelines, tips, and if I am still failing, then the problem is with me, not you.
You need to learn this about who you're mentoring. When they're struggling because they can't learn, and when they're struggling because they need to learn something.
In practice, it's way more complex than this. Looking on "man curl" about how to send headers takes one minute, so to me it looked like an offense when he was telling me how to do. Now if he asked me something harder, that would require me hours of studying, than perhaps he telling me what do would be better, since the problem will be resolved faster.
As a mentor, that's whay you need to learn. You need to balance all these factors to decide if you need to step in or not.
But first of all, like I said in the beginning, trust is the first step. Believe that your student can accomplish something that you thrown at them, and only interfere when the deadline approaches, and the problem needs to be resolved right now. And then later, review what happened. Will this person ever learn this? Is this really that hard that learning about it takes 5 hours instead of 1 hour? Does that mean perhaps he's not the best person for the task? Does this mean that he will never be the right person for the task or will he grok it sometime later?
Questions, questions everywhere ;)
"If you don't create an atmosphere in which risk can be easily taken, in which weird ideas can be floated, then it's likely you're going to be producing work that will look derivative in the marketplace. [...] Those kind of irrational what-ifs eventually lead to something that makes you go, 'Wow, I never would have thought about it.'"
"The university's primary purpose is to build morale, spirit and communication among employees. 'If you could create good filmmakers who would work here for 25 years, their first five years of film would be really good; their next five years would be amazing. By the time these people worked together for 25 years, you would just not believe the things that would happen.'"
Pixar University: Thinking Outside The Mouse
http://www.sfgate.com/cgi-bin/article.cgi?file=/gate/archive...
Also: whacking Ed Catmull with a balloon. Ed I-coinvented-computer-graphics Catmull. Whack on the head with a long thin red baloon.
There seems to be momentum gathering to improve education. Watching Bill Gate's TED video this morning and now this (albeit perhaps more motivational than educational), I'm aware this is a topic which fundamentally interests me but haven't considered yet. It seems an area ripe for innovation in the same way we've seen happen in computing.