Skip to content

Instantly share code, notes, and snippets.

@gtallen1187
Created November 2, 2015 00:02
Star You must be signed in to star a gist
Save gtallen1187/e83ed02eac6cc8d7e185 to your computer and use it in GitHub Desktop.
A little bit of slope makes up for a lot of y-intercept

"A little bit of slope makes up for a lot of y-intercept"

01/13/2012. From a lecture by Professor John Ousterhout at Stanford, class CS140

Here's today's thought for the weekend. A little bit of slope makes up for a lot of Y-intercept.

[Laughter]

So at a mathematical level this is an obvious truism. You know if you have two lines, the red line and the blue line and the red line has a lower Y-intercept but a greater slope then eventually the red line will cross the blue line.

And if the Y-axis is something good, depending on your definition of something good, then I think most people would pick the red trajectory over the blue trajectory (..unless you think you're going to die before you get to the crossing point).

[Laughter]

So in a mathematical sense it's kind of obvious. But I didn't really mean in a mathematical sense, I think this is a pretty good guideline for life also. What I mean is that how fast you learn is a lot more important than how much you know to begin with. So in general I say that people emphasize too much how much they know and not how fast they're learning.

That's good news for all of you people because you're in Stanford and that means you learn really, really fast. This is a great advantage for you. Now let me give you some examples. The first example is: you shouldn't be afraid to try new things even if you're completely clueless about the area you're going into. No need to be afraid about that. As long as you learn fast you'll catch up and you'll be fine.

For example I often hear conversations the first week of class where somebody will be bemoaning, "Oh so-and-so knows blah-blah-blah, how am I ever going to catch up to them?" Well, if you're one of the people who knows blah-blah-blah it's bad news for you because honestly everyone is going to catch up really quickly. Before you know it that advantage is going to be gone and if you aren't learning too you're going to be behind.

Another example is that a lot of people get stuck in ruts in their lives. They realize they're in the wrong job for them. I have the wrong job or the wrong spouse or whatever…

[Laughter]

And they're afraid to go off and try something new. Often they're worried, I'm going to really look bad if I go..

[Laughter]

I'm kidding about the spouse. But, seriously people will be afraid to try some new thing because they're worried they'll look bad or will make a lot of rookie mistakes. But, I say, just go do it and focus on learning.

[Laughter]

Let me take the spouse out of the equation for now.

[Laughter]

Focus on the job.

Another example is hiring. Before I came back to academia a couple of years ago I was out doing startups. What I noticed is that when people hire they are almost always hire based on experience. They're looking for somebody's resume trying to find the person who has already done the job they want them to do three times over. That's basically hiring based on Y-intercept.

Personally I don't think that's a very good way to hire. The people who are doing the same thing over and over again often get burnt out and typically the reason they're doing the same thing over and over again is they've maxed out. They can't do anything more than that. And, in fact, typically what happens when you level off is you level off slightly above your level of competence. So in fact you're not actually doing the current job all that well.

So what I would always hire on is based on aptitude, not on experience. You know, is this person ready to do the job? They may never have done it before and have no experience in this area, but are they a smart person who can figure things out? Are they a quick learner? And I've found that's a much better way to get really effective people.

So I think this is a really interesting concept you can apply in a lot of different ways. And the key thing here I think is that slow and steady is great. You don't have to do anything heroic. You know the difference in slopes doesn't have to be that great if you just every day think about learning a little bit more and getting a little bit better, lots of small steps, its amazing how quickly you can catch up and become a real expert in the field.

I often ask myself: have I learned one new thing today? Now you guys are younger and, you know, your slope is a little bit higher than mine and so you can learn 2 or 3 or 4 new things a day. But if you just think about your slope and don't worry about where you start out you'll end up some place nice.

Ok, that's my weekend thought.

[Applause]

@codingknite
Copy link

This is very nice🔥🔥

@SteveRuben
Copy link

Very nice

@philippe9
Copy link

Interesting

@phongnthai
Copy link

Excellent talk!

@DevComplex
Copy link

This is awesome!

@jasondotparse
Copy link

I wish I had read this years ago.

@hq6
Copy link

hq6 commented Feb 13, 2023

I wonder, is there a primary and/or reliable source to cite this?

@salutis You can probably use this:
https://web.stanford.edu/~ouster/cgi-bin/cs142-spring12/weekendThoughts.php

@MyroslavGryshyn
Copy link

Thanks! Very inspiring!

@frhack
Copy link

frhack commented Feb 13, 2023

If you start early, even with a small slope and a low intercept, you will surpass many with higher intercept but with zero slope.
In other words: start learning now and don't stop !

@rudolf-adamkovic
Copy link

@ianZzzzzz
Copy link

thanks professor

@RubanovDmitriy
Copy link

Thanks! But how anyone can understand if person is a quick learner on interview?

@hq6
Copy link

hq6 commented Apr 26, 2023

I usually look for curiosity and lack of ego as a proxy when I'm interviewing people, because I believe the combination of these attributes helps people learn quickly.

Perhaps @johnousterhout would be willing to answer how he looks for this during interviews?

@johnousterhout
Copy link

Unfortunately I don't have a crisp algorithm for determining how quick a learner a person is during an interview, but I like @h6q's ideas. Some other thoughts: do they have a flexible mind (open to suggestions, quick to consider alternatives or change their mind when wrong), or do they seem rigid (stick to an idea no matter how much evidence piles up against it). If you ask someone to solve a nontrivial problem and talk their way through it, I think you can get some sense of this. But it's not easy... you'll know for sure about 3 months after you've hired them :-)

@romilly
Copy link

romilly commented May 25, 2023

@johnousterhout Yesterday, exploring this theme, I found a short but delightful video clip of you delivering this Friday Thought. Today, search as I might, I cannot find it. Do you happen to know where I might track it down? I'm using the idea for a fun tiny hardware project that illustrates the concept, and I'd like to include a link to the video.

@johnousterhout
Copy link

Unfortunately I'm not aware of any such videos; I'd be curious to see it myself, if you rediscover it :-)

@romilly
Copy link

romilly commented May 25, 2023

It was brief, but it was you, and it was delightful. I will do my best to find it again, and will let you know of I do.

@devvspaces
Copy link

this is an amazing read!

@ShankarAtSochel
Copy link

This is great visualization of "fail fast and learn continuously" - one for all non-math folks out there!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment