I stumbled across this gem from a Stanford lecture by John Ousterhout — creator of Tcl and general computer science luminary — and it’s been rattling around in my head ever since. The quote is simple:
A little bit of slope makes up for a lot of y-intercept.
Now, I know what you’re thinking. “Great, more math metaphors.” But hear me out, because this one is actually profound in its simplicity.
The Math Part (Don’t Worry, It’s Quick)
Picture two lines on a graph. One starts way up high on the y-axis — that’s your y-intercept, your starting advantage. The other starts near zero but has a steeper slope. Given enough time, that steeper line will always overtake the one that started higher.
Always.
It’s not even close after a while.
What This Actually Means
Ousterhout was talking about learning, about how the rate at which you learn is more important than what you know right now. And honestly, this resonates with me in a way that few things do.
When I joined Amazon back in ‘99, I was a UNIX Administrator. That’s it. I didn’t know pricing algorithms, I didn’t know retail, and I certainly didn’t know anything about building the kind of systems that would eventually handle billions in transactions. But I was learning. Fast.
That mispriced iPod incident I wrote about? The one where I built the ASIN Spike Report in a single day? That wasn’t because I knew everything about pricing or anomaly detection going in. It was because I could learn quickly, connect dots, and ship something that solved a real problem. My y-intercept was basically zero in that domain, but my slope was steep enough that it didn’t matter.
The Trap of the High Y-Intercept
Here’s the thing that Ousterhout gets at (and I’ve seen this play out countless times): people with high y-intercepts — lots of initial knowledge, impressive credentials, years of experience — can get complacent. Why learn something new when you already know so much?
But the world moves. Technologies change. Business models evolve. And if your slope is flat, if you’re not learning, you’re falling behind even while standing still.
I’ve seen this in hiring too. The candidate with the perfect resume who hasn’t learned anything new in five years versus the one who’s been teaching themselves new frameworks every few months? I’ll take the learner every time. That slope will make up for the experience gap faster than you’d think.
Asking the Question
Ousterhout’s advice was to ask yourself every day: “Have I learned one new thing today?”
One thing. Not ten. Not a whole new framework or language (though those are great too). Just one thing.
This is basically the same philosophy I wrote about with incremental work — that bird moving the mountain one stone at a time. It’s not sexy. It doesn’t make for good LinkedIn posts about weekend coding marathons. But it’s sustainable, and over years, it compounds into something remarkable.
My Own Slope
Looking back at my career, the common thread isn’t that I knew more than anyone else at any given moment. It’s that I kept learning. Neutrino astrophysics to UNIX administration to pricing algorithms to founding a company. None of those transitions made sense on paper. My y-intercept was laughably low each time.
But the slope? The slope was there.
And here’s what I’ve learned: you can always increase your slope. Always. It’s not fixed. Reading, side projects, asking questions, trying things that might not work — all of these steepen the line.
So What Now?
If you take nothing else from this, take this: don’t be intimidated by people who know more than you right now. Don’t skip opportunities because you feel like you’re starting from zero.
Focus on the slope.
Learn one thing today. Then do it again tomorrow. And the day after that.
Eventually, you’ll look back and realize you’ve traveled farther than you ever thought possible — not because you started with some massive advantage, but because you kept moving.
That’s the power of a little bit of slope.